Diffuse lamellar keratitis (DLK) is a non-infectious diffuse inflammatory reaction that occurs at the interface between the flap and the corneal stromal bed after LASIK surgery. It is characteristically described as “Sands of Sahara,” with fine white granular infiltrates scattered under the flap.
In the Refractive Surgery Guidelines (8th edition), DLK is listed as a postoperative complication for both excimer laser surgery and SMILE, requiring appropriate management 1). The overall incidence in LASIK is reported to be about 0.1-1% 2). Even after femtosecond laser flap creation became mainstream, DLK occurs at a certain frequency, and DLK-like interface inflammation is also observed after SMILE. According to the KLEx (Small Incision Lenticule Extraction) guidelines, the overall incidence of DLK after SMILE is 0.84%, with Grade I (peripheral limited) 1.42%, Grade II (central extension) 0.29%, Grade III (confluent/focal) 0.08%, and Grade IV (severe) 0.02% 3).
The pathology of diffuse lamellar keratitis (DLK) is completely different from infectious keratitis, and the essence of this disease is that bacteria are not involved. On the other hand, distinguishing it from IFS (interface fluid syndrome/PISK), an inflammatory disease with a similar appearance, is the most important clinical challenge because the treatment strategy changes 180 degrees.
QAfter what type of surgery does DLK occur?
A
DLK most commonly occurs after LASIK surgery, but can also occur after SMILE (small incision lenticule extraction). The Refractive Surgery Guidelines (8th edition) list DLK as a postoperative complication for both excimer laser surgery and SMILE. PRK does not create a flap, so DLK does not occur. Even after the widespread use of femtosecond lasers, the incidence of DLK has not been reduced to zero, and meticulous observation and early intervention from 1 to 5 days after surgery are key to a good prognosis. In particular, early detection of Grade 1-2 and prompt initiation of steroid treatment are most important to prevent progression to Grade 3-4. Note that while PRK does not cause DLK because no flap is created, there is a risk of postoperative haze (subepithelial opacity), and prophylactic treatment with mitomycin C may be administered. Individual patient explanations based on a thorough understanding of the surgical technique are required.
LASIK postoperative Stage 3 DLK (diffuse lamellar keratitis): "Sahara sand"-like granular infiltration and resolution after treatment
Lin H-Y, Ho W-T. Diffuse lamellar keratitis as a rare complication of diamond burr superficial keratectomy for recurrent corneal erosion: a case report. BMC Ophthalmol. 2022;22:362. Figure 2. PMCID: PMC9450270. License: CC BY 4.0.
Stage 3 DLK diagnosed 2 days after LASIK: slit-lamp microscopy and external photographs (A) mild corneal edema and ciliary injection, (B–D) “Sahara sand”-like granular white infiltration spreading across the visual axis in the interface under the flap, (E–F) resolution of infiltration after steroid eye drop treatment. Corresponds to the slit-lamp findings (granular infiltration, Grade classification) discussed in the section “Main Symptoms and Clinical Findings.”
Grade 4: Severe decrease in vision (less than 0.5), intense eye pain, photophobia. Marked blurred vision due to stromal melting. Rapid progression may occur from Grade 3 within 24 hours.
The typical onset is 1-5 days after surgery, and it is thought that inflammatory mediators accumulate immediately after flap creation and become manifest. However, late-onset cases several weeks to months after surgery have also been reported, and some cases recur when steroid eye drops are discontinued.
Retroillumination observation with a slit-lamp microscope provides the most important findings for diagnosing DLK. When observed with retroillumination, light comes from behind, making it easier to assess the status of cellular infiltration under the flap. Beginners may find the description “sandstorm-like appearance” intuitively difficult to understand, but with experience, it becomes recognizable as a very characteristic finding. In Grade 1 to 2, granular infiltration is mainly distributed in the periphery, and in Grade 2 or higher, infiltration progresses toward the center. The term “Sands of Sahara” is used as a clinical expression that accurately describes the spread of fine white granular infiltration.
Slit-lamp microscope findings
Granular infiltration: Fine white to gray-white granular infiltration scattered under the flap. “Sand-like” appearance.
Distribution changes: Grade 1 is predominantly peripheral. Grade 2 or higher progresses toward the center.
Border characteristics: The lesion has unclear borders and tends to spread from the flap edge toward the center.
Anterior chamber inflammation: Anterior chamber inflammation is usually mild or absent. If anterior chamber inflammation is severe, differentiation from infectious keratitis should be prioritized.
Anterior segment OCT findings
Hyperreflective areas: Punctate to linear hyperreflective areas are observed at the flap interface.
No flap dehiscence: Unlike IFS, physical dehiscence or fluid accumulation of the flap is not observed (except Grade 4).
Monitoring of stromal melting: In Grade 3 to 4, the melting status of the stromal bed can be evaluated over time using anterior segment OCT.
Differentiation from IFS: In IFS, a uniform fluid layer (hyporeflective zone) is observed under the flap, whereas in DLK, punctate hyperreflective areas are predominant.
Multiple triggers are involved in the onset of DLK. All share a common mechanism: “inflammatory substances are enclosed in the closed space under the flap.”
Endogenous triggers:
Corneal stromal ablation debris and collagen fragments generated during flap creation
Inflammatory cytokines (IL-1, TNF-α, etc.) and lipid breakdown products derived from epithelial cells
Stimulatory substances from plasma formation and microbubbles generated by femtosecond laser
When epithelial tissue from the limbus is entrapped at the incision edge
Exogenous triggers:
Residual metal microparticles from surgical instruments (microkeratome, spatula)
Foreign substances such as glove talc or silicone oil
Corneal epithelial damage from preservatives in preoperative eye drops (e.g., benzalkonium chloride)
Residues of chemicals used in sterilization (glutaraldehyde, hydrogen peroxide, etc.)
Impurities in irrigation solutions or BSS (e.g., endotoxin contamination)
Environmental and facility factors:
Clusters of DLK (multiple cases occurring in a short period within a facility) often suggest contamination in the operating room (organic volatiles, endotoxins, etc.)6). It is important to identify the cause by monitoring the overall DLK incidence rate in the facility and implement environmental improvement measures. Associations with specific lots of cleaning agents, sterilization solutions, or eye drops have been reported.
Infectious inflammation (bacterial, fungal, Acanthamoeba) is excluded by definition of DLK, but infection can trigger a DLK-like reaction.
Factors affecting incidence:
Device: No significant difference in incidence between femtosecond laser and microkeratome2)
Surgeon experience: Higher incidence tends to occur early in the learning curve
Preoperative eye drops: Use of BAK-free formulations may reduce ocular surface inflammation and lower DLK risk
Patient factors: Atopic predisposition and preoperative ocular surface inflammation increase the risk of occurrence.
The most important differential diagnoses for DLK are the following two conditions.
Differentiation from IFS (Interface Fluid Syndrome / PISK):
IFS is fluid accumulation under the flap due to steroid-induced intraocular pressure elevation, without inflammatory cells, and the core issue is elevated intraocular pressure. If DLK is misdiagnosed as IFS and steroid treatment is continued, intraocular pressure rises further, worsening IFS4). Conversely, if IFS is misdiagnosed as DLK and steroids are withheld, DLK may progress to corneal stromal melting.
Differentiating Features
DLK
IFS (PISK)
Onset time
1–5 days postoperatively
Several days to months postoperatively (delayed cases possible)
Infectious keratitis (bacterial, fungal, Acanthamoeba) often presents with anterior chamber inflammation. Stromal inflammation caused by highly virulent bacteria or fungi may show fibrin exudation in the anterior chamber. In non-infectious keratitis (including DLK), anterior chamber inflammation is usually mild. If there are localized infiltrates, hyperemia, and discharge, suspect infection.
The Refractive Surgery Guidelines (8th edition) 1) mandate slit-lamp microscopy on the day after surgery, but for DLK management, the following observation frequency is recommended:
Postoperative day 1 (next day): Examine the interface with slit-lamp microscopy to determine the presence and grade of DLK. If Grade 1–2 is found, start or increase steroid eye drops and instruct the patient to return the next day.
Postoperative days 2–5 (suspected DLK): After starting steroid eye drops, re-evaluate daily or every other day to monitor grade changes. If progression from Grade 2 to 3 is observed, perform flap lift and irrigation immediately.
Postoperative day 7: Confirm improvement of DLK. If Grade 1 or lower, begin gradual tapering of steroid eye drops.
Postoperative day 14 and beyond: Continue intraocular pressure monitoring during steroid tapering, and prepare for transition to IFS.
Most Grade 1–2 DLK resolves completely within 2 weeks postoperatively, but Grade 3–4 may leave permanent irregular astigmatism due to stromal melting, so follow-up with corneal topography for 3–6 months after treatment completion is recommended 9).
QHow do you differentiate DLK from IFS?
A
The most important point in differentiating DLK from IFS (interface fluid syndrome) is intraocular pressure. In IFS, intraocular pressure is elevated, but central Goldmann applanation tonometry may give falsely low readings due to the fluid cushion effect, so peripheral or dynamic contour tonometry is recommended. On anterior segment OCT, DLK shows punctate hyperreflective areas, while IFS shows a uniform hyporeflective fluid layer lifting the flap. DLK presents early (1–5 days) with pain and white granular infiltrates, whereas IFS mainly presents with blurred vision and elevated IOP without inflammatory cells. Because treatments are completely opposite (DLK → increase steroids, IFS → stop steroids + lower IOP), misdiagnosis leads to serious consequences.
Frequent instillation of steroid eye drops is the first-line treatment.
Prednisolone 1% (Predonine ophthalmic solution, etc.): Frequent instillation every 1–2 hours. Assess efficacy within 24 hours. Antibiotic eye drops (fluoroquinolones, etc.) are often prescribed concurrently.
Fluorometholone 0.1–0.5%: Lower risk of IOP elevation than prednisolone, but weaker effect; therefore, switching to prednisolone 1% is recommended for progressive Grade 2 cases.
For Grade I in SMILE, fluorometholone 6–8 times/day is standard management 3).
Re-evaluate the next day to confirm grade improvement. If no improvement in Grade 2, consider flap lift the following day.
During steroid eye drop administration, measure IOP regularly to prepare for transition to IFS. Especially in steroid responders, IOP may rise rapidly; IOP re-measurement 3–5 days after starting treatment is recommended.
For Grade 2, peripheral IOP measurement using Tono-Pen etc., in addition to Goldmann applanation tonometry, is useful to avoid falsely low readings 4).
While continuing steroid eye drops, strongly consider flap lift and saline irrigation.
Flap lift (reopening): Carefully lift the flap using a dedicated spatula. If adhesions remain at the flap edge, bluntly dissect to expose the interface.
Irrigation with BSS (balanced salt solution): Thoroughly irrigate the interface using a 30 mL BSS syringe to physically remove inflammatory cells, debris, and inflammatory mediators such as MMPs.
Flap repositioning: After irrigation, accurately reposition the flap, confirm air bubbles and interface uniformity, then close the eye.
Enhanced postoperative steroid eye drops: After flap lift, continue frequent administration of prednisolone 1% eye drops every 1–2 hours, and reassess the next day.
Eye patch management: After flap lift, consider using a BCL (bandage contact lens) to assist flap re-fixation.
For Grade III in SMILE, consider interface irrigation in addition to high-dose steroids 3). Since interface irrigation after SMILE is more difficult to access than after LASIK, it should be performed by a surgeon with sufficient experience.
Some facilities perform interface injection of dexamethasone 0.1% in addition to BSS irrigation (evidence is limited).
Consider short-term administration of systemic steroids (oral prednisolone 0.5–1 mg/kg/day). Limit the duration to 1–2 weeks and taper to manage side effects.
For Grade IV in SMILE, steroid interface irrigation is mandatory, and systemic steroids should also be added 3).
Since stromal melting often has already progressed in Grade 4, after flap lift and irrigation, monitor the degree of interface and stromal melting over time using anterior segment OCT, and confirm that no residual melting progression remains.
If melting reaches the deep or central area, explain to the patient the future options of corneal transplantation (PTK, keratoplasty, corneal transplant).
Consider flap lift if progression to Grade 2 or higher
2
Same as above, increase dose. Confirm at next day visit
Next day
If no improvement, flap lift the next day
3
Steroid eye drops + flap lift and irrigation
Same day to next day
Flap lift performed in principle
4
Emergency flap lift and irrigation (consider systemic steroids)
Same day emergency
Immediate intervention
QWhat should I do if I am diagnosed with DLK?
A
Treatment for DLK varies by grade. For Grade 1–2, treatment involves frequent instillation of prednisolone 1% ophthalmic solution every 1–2 hours. Most cases improve within 1–2 weeks with this treatment. For Grade 3, in addition to eye drops, flap lift (reopening of the flap) and saline irrigation may be necessary. Grade 4 (stromal melting) requires urgent flap lift and irrigation, and systemic steroids may also be considered. Do not discontinue eye drops on your own; if you experience decreased vision or severe photophobia after surgery, contact your doctor promptly.
DLK is a non-infectious inflammatory reaction primarily involving polymorphonuclear leukocytes (neutrophils). When debris from flap creation, particles from surgical instruments, or biological irritants become trapped in the closed space of the corneal interface, local production of cytokines (IL-1β, IL-6, TNF-α, etc.) and chemokines (IL-8, MCP-1, etc.) occurs, leading to neutrophil accumulation. Unlike infectious inflammation, this reaction does not involve bacterial or fungal proliferation.
Because the flap does not fully re-adhere after surgery, the interface functions as a semi-closed space. This structural characteristic of “containing” inflammatory cells and mediators localizes DLK beneath the flap.
Since the flap interface has limited contact with the normal corneal stroma, the main routes for neutrophil recruitment are thought to be through the limbal vascular network and existing cells within the corneal stroma (keratocytes, Langerhans cells). The speed of progression from onset to severe DLK (Grade 1 to 4) varies greatly among individuals; in some cases, it can rapidly progress to Grade 4 within 1–2 days postoperatively. This is why meticulous observation during the first few days after surgery is absolutely essential for DLK management.
Mechanism of Corneal Stromal Melting (Keratolysis)
In severe DLK (Grade 4), large amounts of matrix metalloproteinases (MMP-8, MMP-9, etc.) derived from neutrophils are released, degrading collagen fibers and causing corneal stromal melting. If melting reaches the central cornea, it can lead to permanent vision loss and irregular astigmatism. Early intervention (flap lift and irrigation) to physically remove MMP-producing neutrophils is a key rationale for preventing progression of melting.
In DLK, pathological findings show accumulation of inflammatory polymorphonuclear leukocytes and mononuclear cells between the layers, whereas in IFS, inflammatory cells are absent and only stromal edema is observed 4). This pathological difference is the fundamental basis for distinguishing treatment strategies between the two conditions.
Femtosecond laser flap creation improves flap accuracy compared to microkeratomes, but laser-induced opaque bubble layer (OBL) formation can cause transient interface reactions. Plasma formation and microbubbles from photodisruption may trigger local inflammatory responses 5). Although a specific increase in DLK associated with femtosecond lasers has not been demonstrated, differentiation between OBL-related inflammation and DLK is considered important.
OBL occurs in 10–30% of cases with femtosecond laser use, and most resolve spontaneously within hours to days postoperatively. However, extensive OBL can interfere with excimer laser eye tracking and reduce ablation accuracy, so if detected intraoperatively, surgeons often wait for resolution. OBL itself is not thought to cause inflammation, but when OBL and DLK are observed simultaneously, retroillumination is particularly useful for assessing DLK progression.
DLK after SMILE occurs in the interface beneath the cap (equivalent to the flap), but the same pathophysiology and treatment principles as LASIK apply 3). However, because instruments are inserted through a 2–3 mm small incision, interface irrigation is more difficult than in LASIK. The KLEx guidelines state that with appropriate steroid treatment, lesions improve within one week in most cases, and symptoms resolve in about three weeks 3).
The incidence of DLK after SMILE (0.84%) is generally similar to that after LASIK (0.1–1%), suggesting that the presence of a flap/cap interface, rather than the specific procedure, is a common trigger. SMILE does not cause flap-related complications (e.g., flap dislocation, free cap), but similar caution is needed in DLK management, and observation on postoperative days 1–5 is particularly important. Interface irrigation for Grade 3–4 DLK after SMILE requires that the surgeon has sufficient experience in postoperative SMILE management 3).
DLK is a complication of refractive surgery for which early diagnosis and appropriate treatment can significantly improve visual outcomes. Establishing a meticulous early postoperative observation schedule and thoroughly instructing patients to seek early care if symptoms worsen are fundamental and most important management strategies to prevent vision loss from DLK.
The incidence of DLK can be reduced by improving the surgical environment. Specifically, quality control of irrigation fluids (purity of distilled water, BSS, etc.), changing surgical gloves (use of talc-free gloves), enhanced ultrasonic cleaning of instruments, and maintaining positive pressure in the operating room to prevent endotoxin contamination are considered important 6). Continuous monitoring of DLK incidence at the facility level is essential for identifying causes and developing countermeasures.
A 2022 review in the Brazilian Archives of Ophthalmology proposed an extended classification that complements the existing Grade 1–4 system, introducing concepts such as fibrotic response (Grade 5) and severe forms with anterior chamber infiltration (equivalent to Grade 6) 7). However, the international consensus currently remains the Grade 1–4 classification as the standard.
Some reports suggest that the use of topical steroids (e.g., prednisolone) before or during surgery may reduce the incidence of DLK 8). However, patients with high steroid sensitivity are at risk of increased intraocular pressure, so individual risk assessment is necessary.
Overview of DLK as a Complication After Refractive Surgery
DLK is a representative condition among complications after refractive surgery that requires diagnostic and therapeutic decisions. A review by Swanson et al. 9) summarized the incidence, risk factors, and treatment outcomes for DLK, epithelial ingrowth, and flap-related complications after refractive surgery, confirming that DLK is the most frequent complication (0.1–1% after LASIK). A review by Phipps et al. 10) systematically organized the latest evidence on the pathophysiology, grade classification, differential diagnosis, and treatment of DLK, emphasizing that conservative treatment leads to recovery in over 90% of Grade 1–2 cases, while early flap lift and irrigation are critical for visual prognosis in Grade 3–4 cases.
A review of flap complications by Moshirfar et al. 11) showed that DLK occurs at similar frequencies with both microkeratomes and femtosecond lasers, arguing that the presence of the flap interface itself is a necessary condition for DLK. The possibility that opaque bubble layer (OBL) formation, unique to femtosecond lasers, may trigger DLK is also discussed 11).
Venkataraman et al. 12) reported a case series of late-onset IFS, including cases that developed more than 10 years after LASIK, demonstrating that IFS can occur at any time after LASIK. They showed that intraocular pressure measurement (using peripheral measurement or dynamic contour tonometry) and anterior segment OCT are essential for differential diagnosis 12).
Customization of Femtosecond Laser and Its Impact on DLK
Slade13) discusses the impact of femtosecond laser flap customization (flap thickness, hinge angle, side cut angle) on flap security and complication profiles, showing that appropriate flap design (uniform thickness, appropriate hinge width, acute side cut angle) contributes to reducing postoperative flap displacement and DLK incidence13).
The ectasia risk scoring system by Randleman et al.14) identified five factors as predictors of postoperative ectasia: abnormal corneal topography, low residual stromal thickness, young age, thin cornea, and high myopia. Although ectasia and DLK are independent complications, when stromal melting occurs in Grade 3–4 DLK, the structural vulnerability of the cornea increases, potentially raising the risk of latent ectasia14).
Corneal Ectasia PPP and Positioning of DLK Management
The AAO Corneal Ectasia PPP15) recommends CXL (corneal cross-linking) as the first-line treatment for managing postoperative ectasia, and indicates that CXL may be considered even in cases of progressive stromal melting due to DLK. However, CXL is indicated only when progression is confirmed; DLK alone is not an indication for CXL15).
Johnson JD, Harissi-Dagher M, Pineda R, et al. Diffuse lamellar keratitis: incidence, associations, outcomes, and a new classification system. J Cataract Refract Surg. 2001;27(10):1560-1566.
Wang Y, Xie L, Yao K, et al. Evidence-Based Guidelines for Keratorefractive Lenticule Extraction Surgery. Ophthalmology. 2025;132:397-419.
Vera-Duarte GR, Guerrero-Becerril J, Müller-Morales CA, et al. Delayed-onset pressure-induced interlamellar stromal keratitis (PISK) and interface epithelial ingrowth 10 years after laser-assisted in situ keratomileusis. Am J Ophthalmol Case Rep. 2023;32:101874.
Kymionis GD, Naoumidi TL, Aslanides IM, Pallikaris IG, Siganos CS. Diffuse lamellar keratitis after laser in situ keratomileusis with the IntraLase femtosecond laser. J Cataract Refract Surg. 2007;33(8):1471-1473.
Stulting RD, Randleman JB, Cowan LA, Thompson KP, Bradley EV, Lynn MJ. The epidemiology of diffuse lamellar keratitis. Cornea. 2004;23(7):680-688.
Rosa DS, de Macedo ELS, Leal LA, et al. Diffuse lamellar keratitis after LASIK: literature review and proposed new grading. Arq Bras Oftalmol. 2022;85(2):198-204.
Venkataraman P, Shroff A, Prabu S, Senthilkumar N. Behind the blur: Understanding interface fluid syndrome in post-LASIK patients. Indian J Ophthalmol. 2025.
Slade SG. The use of the femtosecond laser in the customization of corneal flaps in laser in situ keratomileusis. Curr Opin Ophthalmol. 2007;18(4):314-317.
Randleman JB, Woodward M, Lynn MJ, Stulting RD. Risk assessment for ectasia after corneal refractive surgery. Ophthalmology. 2008;115:37-50.
American Academy of Ophthalmology Corneal/External Disease Preferred Practice Pattern Panel. Corneal Ectasia Preferred Practice Pattern. San Francisco, CA: AAO; 2024.
Copy the article text and paste it into your preferred AI assistant.
Article copied to clipboard
Open an AI assistant below and paste the copied text into the chat box.
