I am writing this to correct a misunderstanding that repeats itself in my clinic every week. If you search for "CXL or CAIRS", most of the pages you find place the two treatments side by side and debate which is "better". That comparison frames the question wrongly from the start.
Cross-linking (CXL, corneal cross-linking) is a treatment in which riboflavin drops and controlled ultraviolet light tighten the collagen fibres of the cornea; its purpose is to halt the progression of the disease. CAIRS (corneal allogenic intrastromal ring segments) changes the geometry of the cornea with thin segments prepared from human donor tissue and placed inside the cornea; its purpose is to reshape for clearer vision. One stabilises, the other reshapes, and the two may not be needed in the same patient at the same time.
In this article I describe the step you may be on: monitoring, CXL, CXL and CAIRS together, and the transplant discussion in advanced disease. I need to draw a line at the outset: this article does not diagnose and does not make a treatment decision. Its purpose is to help you be clear about where you stand when you talk to your doctor. If you are curious about how these treatments have developed, you can read how CXL, CAIRS and transplantation evolved over the last 10 years; my focus here is different: not the history, but where you stand today.
"Which is better" is the wrong question
CXL does the halting, CAIRS does the reshaping: they do not do the same job
This is the point my patients confuse most often, so I separate it at the very beginning:
Halts the disease
Prevents progression and strengthens the cornea. However, it does not bring back vision that has been lost.
Restores vision
It does not halt progression. It targets visual clarity by changing the geometry of the cornea and reducing irregularity.
Setting the two treatments against each other as "which is better" makes as little sense as asking which rung of a ladder is better. The question to ask is: where on that ladder are you?
The limit of this article: it does not diagnose or decide; it clarifies where you stand
The four steps below describe the four most typical pictures I see in clinic. But every patient's cornea is individual; which step you are on only becomes clear with topography, pachymetry and a detailed examination. This article offers a framework for asking the right questions on the way to that examination; it does not replace the examination.
Step 1 — Monitoring only
Who is on this step: stable, non-progressive keratoconus
What my patients on this step have in common is this: the topography findings are consistent with keratoconus, but there is no meaningful change between consecutive measurements. In clinical practice, the framework most often used to describe the stage of the disease is the four-stage classification defined by Krumeich and widely known as "Amsler-Krumeich".1 The current international consensus document approaches staging, follow-up frequency and treatment thresholds with a similar logic.2 But the stage alone does not determine the step; what determines it is whether that stage stays the same over time.
How often is monitoring done, and what is looked at
What I do in stable keratoconus is not to "discharge" the patient but to keep seeing them at regular intervals. Corneal topography, pachymetry (thickness measurement) and visual acuity are repeated at set intervals; what matters is not any single number in itself, but whether those numbers have changed compared with the previous measurement. I will cover how to read the numbers in your report line by line in a separate article.
Not "nothing is being done", but "it is not needed yet"
I say this to my patients explicitly because monitoring often creates a feeling of "I have been neglected". Yet in stable keratoconus, early surgery means adding unnecessary risk to an eye that is not at risk. Monitoring is not a wait outside treatment; it is the form of treatment that is right for that moment.
Clinical threshold note: The current international consensus document sets out updated definitions and recommendations for the diagnosis, staging and follow-up of keratoconus.2 Whether the classification used is Krumeich or Belin ABCD, what matters is that follow-up is carried on consistently with the same system, whichever is chosen.
What this means in practice: If your topography report has not changed over the past year and your doctor recommends monitoring, that is a decision, not a postponement. Not skipping your check-up interval is the only way to catch early whether your step has changed.
Step 2 — CXL only
Who is on this step: there is evidence of progression, and the corneal shape has not yet seriously impaired vision
The moment of moving to this step is when my patients most often ask "why now". The patient here shows a real change on consecutive topographies, such as an increase in maximum keratometry (Kmax), thinning, or a marked shift in refraction. But the corneal geometry has not yet become irregular enough to seriously impair vision; in other words, the real need is to halt, not to reshape.
What CXL aims at, and what it does not
Cross-linking stiffens the cornea by applying riboflavin and controlled ultraviolet-A light to create new bonds between collagen fibres. The goal is to change the course of the disease: to prevent further steepening. CXL is not expected to improve vision markedly; although some patients see a small improvement over time, the real gain is the preservation of vision that would otherwise be lost.
Clinical threshold note: how the decision is made on published progression criteria
I have to be honest here: there is no single, universal numerical threshold for saying "it is progressing". A systematic review published in 2025, covering 221 articles, showed how scattered the definition of progression is in the literature: 85.5% of the articles used maximum keratometry (Kmax) as a criterion, 37.6% used pachymetry and 32.1% used change in refraction; progression was assessed between 6- and 12-month follow-up in 64.7% of the articles.3 So when your doctor says "it is progressing", that is not one number being exceeded; it is the result of several parameters being assessed together, over time.
What this means in practice: If CXL has been recommended, it does not mean your cornea is "bad"; it means the disease is still on the move. Going into the decision knowing that your expectation from the procedure is "not to lose more" rather than "to see more clearly" prevents disappointment later.
Step 3 — CXL + CAIRS together
Who is on this step: there is progression AND the corneal shape clearly impairs vision
The patient on this step carries two conditions together: there is evidence of topographic progression, and the irregularity in corneal geometry now impairs vision to a degree that glasses or standard contact lenses cannot correct. CXL alone does not solve this patient's visual complaint; CAIRS alone does not secure the durability of the gain if progression continues.
Why together: one halts, one reshapes, and neither guarantees the other
Here the two treatments are not alternatives to each other; they complement each other. While CXL aims to strengthen the cornea and halt progression, CAIRS aims at clarity by changing the irregular geometry with tissue segments placed inside the cornea. If only CAIRS is done and CXL is skipped while the disease is still active, the durability of the shape improvement cannot be reliably guaranteed.
The question of sequence: when CAIRS first, when CXL first or simultaneously
There is no single established rule here in the literature or in clinical practice, and I need to say so plainly. In a retrospective study comparing 69 patients treated simultaneously (33 patients) or sequentially (36 patients; CXL first, segment at least 6 months later), the simultaneous group showed a more marked improvement in maximum keratometry (Kmax) and in flat (K1) and steep (K2) keratometry; complication rates did not differ significantly between the two groups.4 In another study, simultaneous accelerated CXL with a ring segment gave more improvement in uncorrected vision than the combination of CXL with topography-guided laser (PRK); the two methods gave similar results in corrected vision.5 These studies show that sequence may affect the outcome, but they are not generalisable enough to say "this should always be done first"; the decision is made according to the thickness of your cornea, the degree of its irregularity and your doctor's experience.
Clinical threshold note: A point of terminology matters: the two studies below were done not with CAIRS prepared from donor tissue but with synthetic ring segments (ICRS). Because the logic of changing corneal geometry by placing a segment is shared, they shed light on the sequencing debate, but we do not generalise their findings directly to CAIRS. In a series comparing simultaneous CXL+ICRS with sequential CXL→ICRS (n=33 vs n=36), a more marked improvement in Kmax/K1/K2 in favour of the simultaneous group and a similar complication rate were reported.4 In another series comparing CXL plus ring segment with CXL plus topography-guided PRK (32 vs 25 eyes), there was a difference in uncorrected vision in favour of the segment group and no significant difference in corrected vision.5 Both studies are single-centre with moderate sample sizes; they should not be read as a generalisable protocol recommendation.
The approach I use: In my own practice I set the sequence according to whether the patient has previously had cross-linking. If no CXL has ever been done, I give priority to CAIRS: I first improve vision and relieve the patient, and after seeing that topography holds the new shape, at least 6 months later, I look at progression and fix that shape with CXL. If the patient comes to me having already had CXL at another centre, that is no obstacle to CAIRS; previous cross-linking does not rule CAIRS out. The literature points the same way: in a study comparing 67 eyes with (30 eyes) and without (37 eyes) prior CXL, the flattening effect of CAIRS was seen in both groups; it was only somewhat more limited in the previously cross-linked eyes (a significant difference in favour of the group without prior CXL in corrected vision and in steep and mean keratometry).9
This is a different sequence from the protocol in which CAIRS was first described: in the first CAIRS series, published by Soosan Jacob in 2018, the segment and cross-linking were performed in the same session (24 eyes, mean follow-up 11.6 months, no progression recorded in any eye during follow-up).10 Over time clinical opinion has diversified on this point; I do not claim that the sequence I prefer is the only right way. The studies above also show that doing both in the same session gives reasonable results.
What this means in practice: If both CXL and CAIRS have been recommended to you, their order is a detail your doctor weighs; doing them in the same session or a few months apart does not mean "the result will be worse". The question to ask is the reason for the sequence, not the sequence itself.
Step 4 — Advanced stage: a segment alone may not be enough
Who is on this step: a very thin, very advanced cornea; the physical limits of segments and CAIRS
In patients on this step the cornea may no longer have the thickness and regularity that allow a segment to be placed; sometimes a scar that has cost the cornea its transparency is added to the picture. What we are discussing here is no longer "which treatment is better" but "what can the tissue physically bear".
A point of honesty: CAIRS does not work at every stage, and why we say so openly
This is the point that many accounts leave out. A meta-analysis pooling 14 studies and 442 eyes reported an improvement of 0.37 logMAR in corrected vision and 0.43 logMAR in uncorrected vision after CAIRS, with an improvement of 4.59 dioptres in refractive error and 4.49 dioptres in maximum keratometry; but this favourable picture belongs to eyes selected as suitable for surgery.6 In a study that compared CAIRS directly with DALK (lamellar transplantation) in advanced keratoconus, corrected vision was significantly better in the transplant group, while uncorrected vision and refractive error were similar. The authors described CAIRS as "an intermediate step before keratoplasty in carefully selected cases".7 So in advanced disease CAIRS is not entirely out of the picture, but neither is it always sufficient on its own.
When lamellar (DALK) and full-thickness transplantation are discussed
Transplantation comes onto the agenda when the cornea has lost its transparency, when excessive thinning threatens surgical safety, or when previous interventions have fallen short. DALK, which we prefer whenever possible, replaces only the diseased anterior and middle layers while preserving the cornea's own posterior layer (the endothelium). In a study of 97 patients comparing 63 full-thickness (PKP) and 39 DALK eyes, postoperative corrected vision was 0.3 logMAR in the DALK group and 0.6 logMAR in the full-thickness group; DALK was reported to carry a lower risk of early graft oedema and rejection.8 For this reason, even when a transplant is needed, we prioritise the lamellar option over full thickness where possible. Transplantation is not the last resort, but it is the last step.
Clinical threshold note: In the CAIRS meta-analysis, the rate of serious adverse events was reported as 0.2% (1 case in 442 eyes, explantation due to graft rejection).6 In the DALK versus CAIRS comparison (26 vs 26 eyes), corrected vision differed significantly in favour of DALK (p=0.027), while uncorrected vision and refractive error were similar.7 These data come from single-centre series of moderate size; in advanced disease, the choice of method is individualised according to corneal thickness, transparency and the patient's priorities.
What this means in practice: Hearing that you are at an advanced stage can be frightening, but it does not automatically mean a transplant. If your corneal thickness and transparency are suitable, CAIRS can still be discussed; and even if a transplant is needed, whether the lamellar (DALK) option is suitable is assessed first.
Where would you place yourself?
The four definitions below are the shortest summary of the steps I have described. Do not place yourself "definitively" in any of them; that only becomes clear with topography and examination. But do see which questions you need to ask.
Monitoring
Topography findings are consistent with keratoconus, but there is no meaningful change between recent measurements.
CXL only
There is evidence of progression; the corneal shape does not yet seriously impair vision.
CXL + CAIRS
There is progression AND the corneal shape clearly impairs vision.
Advanced stage / transplant discussion
The cornea is very thin or has lost its transparency; the physical limits of a segment are being approached.
Who decides the step, and how
You cannot determine which of these steps you are on by yourself, by reading a comparison article online; neither can I, without an examination. The decision is made by weighing three things together: corneal topography and pachymetry, a detailed examination, and above all change over time. I will cover how to read your topography report line by line in a separate article; understanding the Kmax, K2 and ABCD values there strengthens your hand when you have this conversation with your doctor.
Frequently asked questions
Are CXL and CAIRS done at the same time, or one after the other?
It varies. In some cases the two treatments are performed in the same session; in others one is done first and the other is added a few months later. In a retrospective study, simultaneous treatment gave a more marked improvement in maximum keratometry (Kmax) than sequential treatment, with similar complication rates. But that is the finding of a single study; your doctor decides the right order for your cornea.
I was offered only CXL, not CAIRS. Is that bad news?
No. It usually means the shape of your cornea is not yet seriously impairing your vision and the real need is to halt progression; in other words you are on the second step, not the fourth. Whether CAIRS needs to be added is reassessed over time, depending on the quality of your vision and your corneal geometry.
Can CAIRS be done without CXL?
Technically CAIRS can be performed without CXL, but CAIRS on its own has no effect in halting the progression of the disease; it is a treatment that changes the geometry of the cornea and aims at clarity. If the disease is still actively progressing, the shape improvement achieved by CAIRS cannot be reliably guaranteed to last without CXL. For this reason the two treatments are usually planned together in patients with suspected progression.
I was told my keratoconus is mild and no treatment was offered. Is that right?
Yes, it can be, and it is not neglect. If your disease is stable (there is no evidence of progression), the current approach is not to go straight to surgery but to monitor at regular intervals; treatment comes onto the agenda if real progression is detected. It means not that nothing is being done, but that it is not needed at the moment.
Can CAIRS be performed in every patient with keratoconus?
No. The group in which CAIRS gives the best results is cases where the corneal shape is markedly distorted but the tissue is thick enough for surgery. In a very advanced, very thin cornea CAIRS alone may not be enough; lamellar or full-thickness transplantation may come onto the agenda in these situations. In one study, CAIRS in advanced cases was described less as a replacement for transplantation than as an intermediate step that postpones it.
If I am at an advanced stage, will I definitely need a transplant?
No, not necessarily. In advanced cases CAIRS can still be a meaningful option for some patients; transplantation is a step discussed when the cornea has lost its transparency, when excessive thinning threatens surgical safety, or when previous interventions have fallen short. The decision is individualised according to the thickness, transparency and shape of your cornea.
Does my step change over time? If I am being monitored today, could I need CXL tomorrow?
Yes, it can change. Keratoconus is a dynamic disease; a cornea that is stable today can show progression over time, especially at a young age and if triggers such as eye rubbing continue. This is why, in patients being monitored, regular check-ups are the only way to catch early whether the step has changed.
Is there a single right answer to whether CXL or CAIRS should be done first?
No. There is no single established sequencing rule in the literature or in clinical practice; the decision varies with the thickness and shape of your cornea, the state of progression and your doctor's experience. The purpose of this article was in any case not to state a definitive order but to help you be clear about which step you are on.
Let us clarify which step you are on
This decision only becomes clear with corneal topography, pachymetry and a detailed examination. You can make an appointment and we can assess it together.
📞 +90 (312) 275 07 07References
- Krumeich JH, Daniel J, Knülle A. Live-epikeratophakia for keratoconus. J Cataract Refract Surg. 1998;24(4):456-463. PMID 9584238
- Gomes JAP, Hafezi F, Ambrósio R Jr, et al. Global Consensus on Keratoconus and Ectatic Diseases-Edition 2. Cornea. 2026;45(7):888-908. PMID 42228627
- Henriquez MA, Larco C, Izquierdo L Jr. Definition of Progressive Keratoconus: A Systematic Review. Cornea. 2025;44(11):1341-1351. PMID 39671084
- Dal A, Canleblebici M, Kutluksaman B, Erdağ M. Effects and safety of combined corneal collagen crosslinking and intrastromal corneal ring segment treatment in patients with keratoconus: a retrospective study. BMC Ophthalmol. 2024;24(1):473. PMID 39478495
- Cohen E, Ong Tone S, Mimouni M, et al. Comparison of long-term outcomes of simultaneous accelerated corneal crosslinking combined with intracorneal ring segment or topography-guided PRK. J Cataract Refract Surg. 2024;50(4):378-384. PMID 38015419
- Friedrich M, Auffarth GU, Soiberman U, Augustin VA, Khoramnia R, Son HS. Visual and Topographic Outcomes After CAIRS for Keratoconus: Systematic Review and Meta-Analysis. Am J Ophthalmol. 2025;276:81-91. PMID 40157443
- Kara N, Sarsu Kocaaslan E, Görmez GN, et al. Comparison of Visual and Topographic Outcomes Between DALK and CAIRS in Advanced Keratoconus. Cornea. 2026. PMID 41915699
- Aldebasi T, Gangadharan S, Alshammari YS, et al. Comparison of clinical outcomes, complications and patient satisfaction following deep anterior lamellar keratoplasty and penetrating keratoplasty. BMC Ophthalmol. 2024;24(1):501. PMID 39548416
- Yucekul B, Tanriverdi C, Taneri S, Keskin Perk FFN, Karaca Y, Kilic A. Effect of Corneal Allogenic Intrastromal Ring Segment (CAIRS) Implantation Surgery in Patients With Keratoconus According to Prior Corneal Cross-linking Status. J Refract Surg. 2024;40(6):e392-e397. PMID 38848056
- Jacob S, Patel SR, Agarwal A, Ramalingam A, Saijimol AI, Raj JM. Corneal Allogenic Intrastromal Ring Segments (CAIRS) Combined With Corneal Cross-linking for Keratoconus. J Refract Surg. 2018;34(5):296-303. PMID 29738584