FormalPara Key Summary Points

Why carry out this study?

The wound-healing response is critical for a successful corneal laser refractive surgery, particularly surface ablation.

Carboxymethylcellulose and hydroxypropyl guar have been widely used for the treatment of dry eyes.

The study hypothesis was that better protection of the ocular surface using modern topical lubricants may add some benefit to the patient recovery after photorefractive keratectomy (PRK) performed to correct myopia.

What was learned from the study?

Using a topical lubricant containing hyaluronic acid plus hydroxypropyl guar seems to be better than the use of a lubricant with carboxymethylcellulose in terms of eye discomfort and speed of corneal re-epithelization after PRK.

Little attention has been paid to the role of topical lubricants in the early postoperative period of PRK. It seems that appropriate lubrication has a major role to play in the postoperative management of PRK.

Introduction

Photorefractive keratectomy (PRK) was the first method in modern refractive surgery to treat myopia, astigmatism, and hyperopia [1]. Healing of the corneal epithelial defect after PRK typically occurs over approximately 5 days [2]. Although PRK is the least invasive refractive method, complete corneal re-epithelization and tear stabilization have clinical relevance as they may affect visual acuity and quality of vision and are related to postoperative patient discomfort [3,4,5,6,7].

Numerous artificial tears are used after PRK to help enhance epithelial regeneration which facilitates early visual recovery and ultimately improves visual outcomes [8, 9]. This is the most commonly used treatment on the ocular surface and has led to the appearance of more advanced components each year, such as carboxymethylcellulose or carmellose sodium (CMC), hydroxypropyl methylcellulose, mineral oils and phospholipids, hyaluronic acid (HA), and hydroxypropyl guar (HP-guar) [10,11,12,13]. These agents can stay on the ocular surface and create a nurturing environment to promote epithelial cell healing. Moreover, they increase the viscosity and stability of the tears, to improve discomfort symptoms and reduce inflammation.

HP-guar is a muco-mimetic gellable agent, with rheological properties very similar to those of tears [14]. HA, a mucopolysaccharide, is the key component of the extracellular matrix, plays a critical role in cell proliferation, anti-inflammation, and wound repair, and has viscoelastic and hygroscopic properties [14]. The efficacy of the HA and HP-guar combination (HA + HP-guar) in the treatment of dry eye symptoms has already been reported for both the perioperative and postoperative management of cataracts [15, 16].

CMC on the other hand is a high molecular weight polysaccharide derived from natural cellulose obtained through chemical modification. CMC binds directly to corneal epithelial cells for several hours through interaction between glucopyranose subunits and the glucose transporter. In addition, its property of binding to matrix protein stimulates the migration of epithelial cells which contributes to the re-epithelialization and healing of corneal wounds [17, 18]. CMC-containing eye drops are effective in the treatment of dry eye disease (DED) [10, 13].

In this study, we compared the effect of two preservative-free (PF) artificial tears, one containing CMC vs another containing HA + HP-guar, on the healing of the corneal epithelium and the ocular discomfort after bilateral PRK surgery.

Methods

This randomized, prospective interventional study was conducted at the Clinica Novovision Madrid, Spain. The study was approved by the local ethics committee (CEIm Regional de la Comunidad de Madrid, Code RC-06/20) and followed the tenets of the Declaration of Helsinki of 1964 and its later amendments. Informed written consent was obtained from all patients for both the participation in the study and the publication of the results.

Patients with myopia who were candidates for PRK to correct myopia in both eyes, and with no history of previous ocular surgery, systemic autoimmune diseases, or under any topical or systemic treatment that could affect the ocular surface, were candidates for the study. Patients with preoperative moderate to severe DED (corneal staining higher than Oxford grade I) were excluded from the study. Patients were advised to stop wearing contact lenses at least 1 week before entering the study (baseline visit).

Sixty-eight patients were treated with PRK surgery performed with the WaveLight® Allegretto Wave® Eye-Q 400 Hz Laser (Alcon Laboratories, Fort Worth Tx, USA). All surgeries were performed by the same experienced surgeon (MAT), at the same center, and with the same preoperative, intraoperative, and postoperative protocols (except for the topical lubricant used). All patients were told to use oral analgesics (metamizole 575 mg) as needed. No topical analgesic was prescribed. All the patients were told not to use make-up till complete re-epithelization, and to use lid-wipes for lid cleaning, if necessary.

The randomization was done using a computer-generated randomized list. The lubricant corresponding to each patient was only known by the optometrist in charge of the randomization process (MC). All other investigators performing the clinical evaluations and the instrumental procedures (tear osmolarity, non-invasive Keratograph tear break up time [NIKBUT]) were masked for the lubricant used by the patient.

Preoperative and postoperative evaluations included slit lamp examination with staining, corneal Scheimpflug tomography, uncorrected and corrected distance visual acuities (UDVA and CDVA, in decimal notation), NIKBUT (Keratograph, Oculus, Wetzlar, Germany), and tear osmolarity (Tear Lab, Escondido, CA, USA) were performed. The NIKBUT has greater discriminative ability in detecting dry eye than the conventional tear breakup time (TBUT) using fluorescein [19]. In addition, the Tear Lab is the preferred method for tear osmolarity measurement in clinical practice because it needs a small volume of tears (50 nL) and has a high precision (coefficient of variation of 2.2%) and a small standard deviation (6.6 mOsm/L) in repeated within-subject measurements [20]. If the results of tear osmolarity, NIKBUT, and slit lamp examination were symmetrical for both eyes, the selected value was the one from the right eye. In case this did not happen, the eye with the worst value was selected. These tests were used at different time points during the follow-up visits (Table 1).

Table 1 Preoperative and postoperative examination protocol
Full size table

Patients were randomized into two groups. In the study group, 34 patients were treated with Systane Ultra Plus Hydration® (sodium hyaluronate 0.15%, polyethylene glycol 400 0.4%, propylene glycol 0.3%, hydroxypropyl guar 0.175%, Alcon Laboratories, Fort Worth, TX, USA). In the control group, 34 patients were treated with ViscoFresh® (CMC sodium 0. 5%; Allergan Inc., Irvine, CA, USA). The recommended dosage was at least four times a day for the first postoperative month and at least twice a day thereafter. Patients were asked to record the actual number of ocular lubricant usage per day (like a patient diary) and to report it during study visits. In case any patient was using another type of ocular lubricant before the study, they were instructed to use only the designated product after surgery.

The ocular surface was evaluated on days 1, 4, 7, 30, and 90 postoperatively, with a slit lamp by the same, masked ophthalmologist. On day 1, measurements of the corneal defect (highest diameter of the epithelial defect) were performed without fluorescein staining, only with a slit lamp, as the patients still had the therapeutic contact lens on. The same procedure was performed on day 4; and in case the contact lens could be removed, measurements were performed with staining to evaluate corneal punctate staining. Measurements on days 7, 30, and 90 were evaluated with staining as well.

The visual analog scale (VAS) was administered after the surgery on day 1. The patient took another VAS test to assess the pain on day 3 and this test was handed in on day 4. From day 4 to day 90, patients were asked to complete the VAS test during their visits (Fig. 1). The aim was to reduce the possible bias induced by the ability of the patient to compare the results from prior tests and consequently induce artifacts in the data.

Fig. 1
figure 1

Visual analog scale (VAS) pain evaluation was performed by the patient at every postoperative visit, including day 2 and 3 (remote)

Full size image

The ocular surface disease index (OSDI) questionnaire was completed by the patients preoperatively and on days 30 and 90, as it evaluated symptoms during a week, which does not validate its use on days 1 and 4, postoperatively (Table 1).

Additionally, on postoperative days 30 and 90, the patients were given a VAS-like questionnaire (scale from 1 to 10) to evaluate how bothered they were by the blurriness caused by the artificial tear application and how long that blurriness lasted (Fig. 2).

Fig. 2
figure 2

Visual analog scale (VAS) to evaluate how bothered patients were by blurriness caused by the artificial tear application and how long that blurriness lasted

Full size image

The order in which the different examinations were performed is shown in Fig. 3.

Fig. 3
figure 3

Flowchart showing the order in which the different examinations were performed. NIKBUT non-invasive Keratograph breakup time

Full size image

Statistical analysis was performed using SPSS version 25. Continuous data were described as mean, standard deviation, or the median, mode, and the interquartile range, and categorical variables as percentages. The unpaired two-tail t test or the nonparametric Mann–Whitney U test were used for comparisons of continuous variables. Categorical variables were compared using the chi-square test. A p value of less than 0.05 was considered to show a statistically significant result.

Results

A total of 68 consecutive patients scheduled for PRK surgery to correct myopia were randomized into two groups. The mean age of the control and study groups was 28.8 ± 5.3 years and 29 ± 5.6 years respectively (p = 0.9). Both groups were comparable in terms of gender and preoperative refractive error magnitude. Preoperative patient demographics are presented in Table 2.

Table 2 Demographics and preoperative examination results
Full size table

No patient reported poor tolerance or any adverse event related to the topical lubricants.

Table 3 presents the postoperative diameter of the de-epithelized cornea, the fluorescein staining using the Oxford scale, and the percentage of eyes showing complete re-epithelization at 1, 4, 7, 30, and 90 days after PRK. Four days postoperatively, the de-epithelized area was significantly smaller in the study group, and the percentage of eyes with complete re-epithelization was also significantly higher in the study group. Both groups demonstrated full re-epithelization on day 7. Fluorescein staining (Oxford scale) showed small values after the first postoperative week in both groups.

Table 3 Postoperative ocular surface parameters (34 patients in each group)
Full size table

Postoperative pain reached its maximum 3 days after PRK. Patients treated with HA + HP-guar reported statistically significant lower pain on day 3, while no statistically significant pain difference between groups was found 1, 4, 7, 30, and 90 days postoperatively (Table 3). Concerning blurred vision after installation of the topical lubricants, VAS showed higher distortion in the HA + HP-guar group 30 days postoperatively but the results were comparable at postoperative day 90 (Table 4).

Table 4 Postoperative ocular pain and blurred vision parameters (34 patients per group)
Full size table

Both groups had no differences in the visual or refractive outcomes, postoperatively. UDVA was above 0.75 on day 7 and above 0.95 on day 30 while full vision recovery was achieved 3 months postoperatively (Table 5). Moreover, there were no differences between groups in the NIKBUT, tear osmolarity, or OSDI score at any time point (Table 5).

Table 5 Postoperative visual and refractive outcomes, tear stability, tear osmolarity, and OSDI values
Full size table

Discussion

The wound healing process after PRK plays a crucial role in the outcome of surgery and patient comfort. Great attention has been drawn to understanding the physiology of the corneal recovery after PRK [21]. PRK surgery entails debridement of the epithelial layer and stromal excimer laser ablation, resulting in damage of the subepithelial nerve plexus and reduced corneal sensitivity, and is commonly associated with the development of dry eyes [7, 8].

Artificial tear drops have been the most common treatment for patients with dry eye symptoms. Although a genuine substitute for human natural tears has not been produced, alternatives with pioneering substances have been successful in alleviating eye discomfort related to dry eye, postoperatively.

In the present study two commercial artificial tear solutions were used for comparison and considered highly effective in alleviating the symptoms of ocular surface disorders [10, 13,14,15,16]. Systane Ultra Plus Hydration, which contains HA + HP-guar, acts through a unique biphasic mechanism of action, in which the product first binds to damaged hydrophobic areas on the ocular surface to add volume to the tear film and then restructures the tear film by forming a protective gel matrix that provides long-lasting protection [22]. ViscoFresh with CMC as the main ingredient increases the viscosity of tears and has mucoadhesive properties that allow it to remain on the ocular surface for long periods [23]. Direct comparisons of these two formulations have been made only in dry eye treatment and not after photorefractive keratectomy [24, 25].

The current study showed that epithelium heals faster using artificial tears with HA + HP-guar compared with conventional PF CMC artificial tears while managing postoperative dry eye. De-epithelized area was significantly lower in the study group than in the control group 4 days postoperatively. However, 1 week postoperatively and thereafter, there was no statistically significant difference in de-epithelized area between both groups. The faster epithelialization in the study HA + HP-guar group was associated with less induced ocular pain and discomfort on day 3. Before and after that time point, the ocular pain and discomfort were comparable among groups. The aforementioned results are in agreement with Christensen et al., who reported that the artificial tear used in our study containing HA + HP-guar was more effective in reducing corneal staining compared to the CMC-containing lubricant, on the 14th and 42nd follow-up day when used for the treatment of the dry eye [24]. Similarly, Maharana et al. reported that the improvement with HA + HP-guar as a tear substitute was consistent at all follow-ups (1 and 4 weeks) for OSDI, tears breakup time, and Schirmer test, compared to CMC 0.5%, concluding that HA + HP-guar can lead to a faster and more sustained relief in both objective and subjective parameters compared to CMC 0.5% tear substitute [25].

The observed differences in the efficacy of the two artificial tear solutions could be attributed to a variety of factors. Some of the ingredients of both formulations are commonly found in over-the-counter tear supplements. The HA + HP-guar solution contains polyethylene glycol  00 (0.4%) and propylene glycol (0.3%) while the CMC solution contains carboxymethylcellulose (0.5%). The difference in viscosity and mechanism of action between the two products might account for some of the observed effects. A unique attribute of HA + HP-guar is the addition of HP-guar as a gelling agent, which is pH sensitive. When exposed to the ocular pH (approximately 7.5), the viscosity increases significantly [24]. A direct comparison of formulations with more similar viscosities would help elucidate the role that HP-guar may play in the formulation.

Moreover, HA + HP-guar solution has a higher molecular weight (1000–5000 kDa) than the cellulose solution (318–1178 kDa) giving different rheological properties. The high molecular weight HP-guar molecules, along with the natural mucin layer, would tend to prolong the retention of the demulcents on the eye [26]. Once this mucin-guar demulcent network has formed over the surface of the eye, it provides additional lubricity to the eye and serves as a temporary “bandage” to allow the natural repair processes of the surface epithelial cell to occur. Maharana et al. found that patients in the HA + HP-guar group had a significantly lower mean OSDI score than those in the CMC group at weeks 1 and 4 after treatment initiation for dry eye [25].

The UDVA, CDVA, subjective refraction, NIKBUT, and osmolarity presented similar results at all time points in both groups. On the other hand, the blurriness in vision caused immediately after topical application of the artificial tears lasted significantly longer in the study group. This difference is probably caused by the increased viscosity of HA + HP-guar solution (10 cp) compared to the viscosity of CMC solution (3 cp). Our findings could be explained by the study of Wang et al., reporting that central tear film thickness, tear volume, and tear meniscus increased just after installation and values are generally proportional to artificial tear viscosity. However, no effect was seen 20 min after the drop installation [27]. This observation was also confirmed by Zaki et al. and Wilson, who found that increasing the viscosity of instilled fluid slowed the fluid clearance [28, 29].

Limitations of the current study included the short duration, which was insufficient to evaluate the impact of the artificial tear treatment on long-term dry eye discomfort after PRK surgery, and the lack of a placebo ophthalmic solution. Other possible limitations are that the patients were myopic, young, and had no previous ocular surgery, so it could be that the results of the current study may not apply to PRK performed in other populations, such as PRK performed for refractive adjustment after cataract surgery.

Conclusion

The current study showed faster healing of the corneal epithelium and a significant reduction of the early postoperative pain and discomfort when using the combination HA + HP-guar (Systane Hydration®) as compared with conventional artificial tears (ViscoFresh®) after PRK. It seems that the best results obtained using Systane Ultra Plus Hydration® are not related to an improvement in the stability or osmolarity of the tear film, but probably to a direct effect on the epithelial cells/superficial stroma/corneal nerves.