Development of interface haze after femtosecond laser-assisted in situ keratomileusis with accelerated corneal crosslinking: a case series

Femtosecond laser-assisted in situ keratomileusis (FS-LASIK) with accelerated corneal crosslinking (FS-LASIK Xtra) is a recent procedure to achieve safer corneal ablation in myopic patients with borderline corneal thickness. Despite its well-accepted effectiveness, the development of remarkable interface haze is a potential concern but has rarely been reported and discussed. We report for the first time a case series of 11 eyes of 7 patients who developed typical interface haze 1–3 months after FS-LASIK Xtra for the correction of myopia with astigmatism, with intensity grades ranging from 0.5 + to 3 + at the time of onset. The preclinical spherical diopters of the 7 patients ranged from − 2.25 D to − 9.25 D and cylindrical diopters ranged from − 0.25 D to − 2.50 D. The haze tended to be self-limiting, and topical anti-inflammatory therapy was given to moderate and severe cases, who responded well to treatment. The development of clinically significant interface haze is a relatively rare complication after FS-LASIK Xtra but tends to have a higher incidence and intensity compared to conventional stromal surgery such as FS-LASIK. Timely treatment and close follow-up are essential to patients undertaking FS-LASIK Xtra.


Introduction
Femtosecond laser-assisted in situ keratomileusis (FS-LASIK) with accelerated corneal crosslinking (FS-LASIK Xtra) is a combination procedure to achieve safe and effective correction of moderate and high myopia with borderline corneal thickness [1]. Both the creation of the ultra-thin flap and the application of corneal crosslinking are intended to preserve and strengthen the biomechanical properties of the remaining corneal stromal bed and reduce the risk of post-operational corneal ectasia [2,3]. Previous studies demonstrated comparable effectiveness between the conventional procedure and the combination method with accelerated corneal crosslinking [4][5][6][7]. However, the safety issue regarding the development of interface haze has been rarely reported and discussed [8,9].
Since the introduction of FS-LASIK Xtra in our refractive surgery center, we noticed a series of cases developing clinically significant interface haze, which surpassed the intensities and prevalence of haze observed in conventional FS-LASIK patients.
Here we report a series of cases of patients developing interface haze after FS-LASIK Xtra in our center and discuss the treatment, prevention, and potential pathogenesis associated with the more pronounced haze formation.

Materials and methods
A total of 58 eyes of 36 patients underwent FS-LASIK with simultaneous accelerated corneal crosslinking (FS-LASIK Xtra) in our corneal refractive center from April to July, 2022, and 11 eyes of 7 patients developed clinically significant interface haze. Their demographic information and pre-operational eye examination results are demonstrated in Table 1. Their preoperational spherical diopters (sph) ranged from − 2.25 D to − 9.25 D and cylindrical diopters (cyl) ranged from − 0.25 D to − 2.50 D. All patients denied a history of previous eye surgeries, eye diseases, or systemic disorders including metabolic diseases, tumors, and immune-associated diseases that could affect corneal healing. In our refractive surgery center, the indications for FS-LASIK Xtra were as follows: (1) patients aged no less than 18 years old; (2) stable refractive status (annual change of refraction ≤ 0.50 D in the past 2 years); (3) normal corneal topography and opacity; (4) patients at higher risk of post-operational ectasia were suggested to take FS-LASIK Xtra instead of conventional FS-LASIK, including the calculated remaining stromal bed thickness after conventional FS-LASIK at the borderline level (280 μm) or young patients aged 18-21 years old with a spherical diopter greater than − 8.00 D; (5) no systemic diseases that could impact post-operational recovery; (6) no other ocular disorders that may compromise the safety or effect of surgery; and (7) patients have the willingness to take FS-LASIK Xtra and have reasonable expectations for the outcomes of the surgery. FS-LASIK Xtra as the most suitable surgical method was introduced to patients, with its risks and benefits clearly explained before surgery. Written informed consent for surgery and using figures and data for publication were also obtained. Patients undertaking FS-LASIK without   accelerated corneal crosslinking during the same period in our refractive center were used as the control. All surgeries were conducted by the same surgeon under topical anesthesia in a sterile operation room following the standard established protocol. Briefly, a 110-μm corneal flap was created with the femtosecond laser (VisuMax ® , Carl Zeiss ® , Oberkochen, Germany) and lifted with a spatula. The stromal bed ablated with the excimer laser (SCHWIND AMARIS ® , Kleinostheim, Germany) following the standard FS-LASIK procedure. Then the stromal bed was soaked with riboflavin sodium phosphate solution (concentration ≥ 0.23%) for 60 s and rinsed with the balanced salt solution. Then the corneal flap was gently restored. Ultraviolet-A (UVA) was applied for 60 s to induce corneal crosslinking, with a power of 30 mW/cm 2 and the total energy delivered as 1.8 J/ cm 2 . Standard topical therapy was given to surgical eyes, including antibiotic eye drops, steroids, and artificial tears.
Patients were regularly followed up in clinics. At each visit, patients underwent a thorough ocular examination, including uncorrected distance visual acuity (UDVA), corrected distance visual acuity (CDVA), optometry examination, intraocular pressure (with non-contact tonometer, Oculus ® , Wetzlar, Germany), and slit lamp examination (Biobase ® , Suzhou, China). In addition, anterior segment optical coherence tomography (AS-OCT, Heidelberg ® , Heidelberg, Germany), corneal topography (Oculus ® , Wetzlar, Germany), and corneal densitometry (Oculus ® , Wetzlar, Germany) were prescribed as needed. The severity of haze was graded subjectively based on the Fantes classification system, with 0.5 + to 4 + representing the mildest to most severe opacities [10]. Table 2 summarizes the details of vision, optometry results, haze intensity, management, and response to treatment of each patient. Of these 7 patients, 4 patients received FS-LASIK Xtra in both eyes, and 3 patients received FS-LASIK Xtra in 1 eye and conventional FS-LASIK in the other eye. Clinically significant haze was observed in all surgical eyes undertaking FS-LASIK Xtra in these patients. The onset time of haze ranged from 1 to 3 months after surgery. The grade of haze ranged from 0.5 + to 3 + at the onset. Topical steroid eye drops were prescribed to 5 patients. Except for patient 4 and patient 5 who developed haze at the latest follow-up visit and no further observation results were available, other patients all showed improvement in haze compared to the onset time. Among the 7 patients, patient 1 developed the most serious haze and complained of a dramatic decline in vision. FS-LASIK Xtra was successfully conducted and the patient had a good UDVA 1 week after surgery (OD 0.8 and OS 1.0 −2 ) and the optometry test of the right eye was + 0.50 sph/− 0.75 cyl × 69° and the left eye was + 0.75 sph/− 1.00 cyl × 92°. Figure 1 demonstrates the reticular stromal clouding of his both eyes at the time of onset, which was 3 months after surgery. As shown in the AS-OCT, the clouding was present right beneath the surgical layer of the anterior stroma. The UDVAs of his right eye and left eye declined to 0.6 +3 and 0.8, which could be corrected to 0.8 +3 and 1.0, respectively. The optometry test of the right eye was − 1.50 sph/− 0.50 cyl × 89°, and the left eye was − 0.50 sph/− 1.25 cyl × 100°. Densitometry of the anterior 120 μm cornea showed a significant increase from the pre-surgical level of 18.8 to 42.2 in the right eye and from 17.1 to 55.3 in the left eye. The patient was given topical intense steroid therapy, including tobramycin/dexamethasone ointment once every night and prednisolone acetate eye drops every 2 h. The patient responded well to treatment and the stromal clouding of both eyes dramatically improved in 1 week, with densitometry values declining to 25.3 and 23.3 in his right and left eyes, respectively (Fig. 2). Then steroid therapy was gradually tapered and the stromal haze continued to improve. After 2 months of treatment, only slight clouding remained at the corneal center and the densitometry values declined to 17.9 and 18.8, respectively (Fig. 3). However, despite the improvement of haze, the UDVA of his right and left eyes worsened and declined to 0.3 and 0.6 at the last visit, which could be corrected to 0.8 and 1.0. Optometry test revealed the development of significant astigmatism in both eyes, which was + 0.50 sph/− 2.50 cyl × 85° in the right eye and 0.00 sph/− 1.25 cyl × 90° in the left eye, respectively. The corneal topography test showed a remarkable change of curvature predominantly in the anterior corneal surface. The mean radius of curvature (Rm) of the anterior surface was 9.63 mm 1 month after surgery and decreased to 9.52 mm at the onset of haze, which then increased to 9.87 mm at the last visit in the right eye. For the left eye, the Rms were 9.67 mm, 9.31 mm, and 9.44 mm at the corresponding visit time, respectively. The patient lost 1 line of CDVA in his right eye compared to the pre-operational level.

Results
During the same period, no patients undertaking conventional FS-LASIK without accelerated corneal crosslinking developed interface haze after surgery. Figure 4 demonstrates the follow-up results of a patients taking FS-LASIK at 1 month after surgery, showing clear cornea and no haze formation.

Discussion
Our case series reported the development of interface haze in 11 eyes of 7 patients undertaking FS-LASIK Xtra to correct myopia in our refractive surgery center. These patients generally had moderate to   high myopia without any previous history of eye diseases or systemic disorders. Both the incidence and intensity of haze seem to surpass the haze reported in patients receiving conventional FS-LASIK. To our best knowledge, this is the first study that reports concerns over the increased risk of haze formation associated with FS-LASIK Xtra.
Haze is most commonly seen in surface ablation surgeries, such as photorefractive keratectomy (PRK). The formation of haze is a complex pathological process, and the interaction between corneal epithelial cells and keratocytes via a cascade of cytokines, growth factors, and chemokines plays a vital role in its formation [11]. During this process, keratocytes get activated and turn into myofibroblasts, leading to the deposition of disordered collagen fibers in the extracellular matrix with the induction signals secreted by corneal epithelial cells, such as tumor growth factor-β (TGF-β) [12]. The intact basement membrane of corneal epithelium acts as a molecular   barrier that restricts the crosstalk between epithelium and stroma; thus, the incidence of haze in conventional corneal stromal surgeries like sub-Bowman keratomileusis, small incision lenticule extraction, and FS-LASIK is greatly reduced [13].
Our case series demonstrated a relatively high incidence of haze in patients undertaking FS-LASIK Xtra in our refractive surgery center. In particular, patient 1 developed pronounced haze in his both eyes, which was rarely reported for corneal stromal surgeries. Currently, there is little and controversial evidence regarding the application of accelerated crosslinking on the formation of haze after corneal stromal refractive surgeries. Corneal crosslinking is a recent technique that applies riboflavin and ultraviolet-A (UVA) to produce reactive oxygen species and result in the formation of chemical bonds between corneal stromal lamella and strengthen its biomechanical properties [14]. Studies in keratoconus patients undertaking corneal crosslinking revealed the incidence of haze formation was as high as over 50% [15]. Histological studies in rabbit corneas revealed that the application of UVA induced the transient induction of fibroblast instead of persistent myofibroblasts transition from keratocytes during conventional haze formation [16]. However, the haze associated with corneal crosslinking appears as dust-like tiny opacities or the formation of a demarcation line in the corneal stroma. In comparison, in our case series, patients presented with reticulate haze, which was located right at the interface level as shown in the AS-OCT. Thus, we believe it should not be fully explained by the crosslinking process or there might be even totally different pathological mechanisms. Previous comparison studies also demonstrated that generally there was no increased risk of haze formation between LASIK and LASIK Xtra [9,17]. However, these studies were limited by a small sample size and lack of objective evaluation methods such as corneal densitometry, and further histological or biological evidence is needed.
In patients with borderline corneal thickness, an ultra-thin flap is created to preserve a thick stromal bed and achieve better post-operational biomechanical strength. However, with a thinner flap, the risk of damage to the basement membrane increases, which enables direct crosstalk between corneal epithelium and stroma and can potentially lead to the formation of haze [18]. Rocha et al. found the risk of interface haze formation after LASIK surgery was strongly associated with the ultra-thin flap setting with a total flap thickness < 90 μm. A previous case series reported 10 patients undertaking LASIK surgery to correct myopia and astigmatism who developed subepithelial mild haze with consequent myopia regression. The estimated flap thickness was only about 70 μm during the operation [19]. To avoid the potential complications associated with an ultra-thin flap, we set the flap thickness as 110 μm, which is much higher than the generally accepted threshold of 90 μm that can be associated with a higher risk of haze formation. However, although the basement membrane barrier remains intact to prevent the crosstalk between the epithelial cells and keratocytes, the application of UVA radiation during FS-LASIK Xtra could promote keratocytes activation and myofibroblast transition directly, which might account for the increased risk of haze formation [20,21].
All patients in our case series responded well to treatment and showed a trend of improvement. The interface haze after FS-LASIK Xtra is selflimited, just like the transitory haze observed after conventional corneal refractive surgery, which may spontaneously resolve in mild cases [22]. For patients with moderate and severe interface haze, topical anti-inflammatory eye drops based on steroids were prescribed. In particular, patient 1 with the most serious haze responded quite well after 1-week intense steroid therapy and prolonged maintenance therapy by anti-inflammatory eye drops. Topical steroid therapy is able to modulate keratocyte activation and inhibit the secretion of TGF-β, a key signaling molecule for haze formation [23]. However, previous clinical studies provide controversial evidence regarding the application of topical anti-inflammatory therapy in the prevention and treatment of haze after corneal refractive surgery [24][25][26][27][28]. In our center, cortical steroid eye drops are conventionally used for 30 days to inhibit inflammation and modulate corneal stromal remodeling after both FS-LASIK and FS-LASIK Xtra. Based on the experience of the 7 patients with haze formation, we have extended the duration of topical steroid therapy for patients receiving FS-LASIK Xtra and noticed a decline of haze formation afterward. However, the benefits and risks of using prolonged antiinflammatory therapy on such patients need further evaluation by well-designed studies in the future. In addition, due to the limited short observation time, the risk of late-onset haze formation associated with FS-LASIK Xtra is not evaluated and is also of great research interest, which is considered as a limitation of our study.
Above all, our case series reported for the first time the higher concerns of haze formation in patients undertaking FS-LASIK Xtra surgery. It is hypothesized the application of simultaneous accelerated crosslinking accounts for the haze formation after FS-LASIK via direct activation of keratocytes, and further explorations are needed. In addition, adequate anti-inflammatory therapy may be a feasible way to reduce haze formation in such high-risk situations.