1. Ocular surface and eyelid observations

In both white and blue light examinations, CFA-injected rats showed successful induction of MGD, as evidenced by substantial eyelid edema when compared to control animals. During the experimental period, eyelid swelling progressively worsened in the untreated MGD group, whereas treated groups—including those receiving LED irradiation, hyaluronic acid, cyclosporine, or thermal therapy—showed visibly reduced edema. On the 22nd day, no substantial corneal ring irregularities were observed in any of the groups, indicating the absence of advanced corneal surface damage (Fig. 3).

Fig. 3

Photographs of MGD-induced rat eyes during the experimental period. MGD-induced groups groups(MGD, MGD+L, MGD+HA, MGD+R, MGD+L+H) have swollen eyelids throughout the experiment, with minimal damage on the corneal surface. A: Surface of rat cornea after inducing alkali burn. B: Eyelid observation during the experimental period. CON: control; CON+LED: control with LED treatment; MGD: meibomian gland dysfunction; MGD+LED: meibomian gland dysfunction with LED treatment; MGD+HA: meibomian gland dysfunction with hyaluronic acid; MGD+R: meibomian gland dysfunction with cyclosporine treatment; MGD+H: meibomian gland dysfunction with thermal treatment; MGD+L+H: meibomian gland dysfunction with LED and thermal treatment

2. Tear film break-up time (TBUT)

TBUT was measured on days 8, 15, and 22. It was observed that all treatment groups exhibited enhancements in comparison to the untreated MGD group. As demonstrated in Fig. 4 A, by the 15th day, TBUT exhibited a substantial increase in the majority of the treated groups, with the exception of the MGD + HA group. On the 22nd day, the MGD + H group demonstrated the most significant enhancement, exhibiting TBUT values that were statistically indistinguishable from those of the normal control group.

Notably, the MGD + LED (p = 0.043) and MGD + L + H (p = 0.004) groups exhibited TBUT enhancements that were analogous to those observed in the MGD + R group (p = 0.000), indicating comparable therapeutic efficacy. The study revealed no statistically significant differences between the CON and CON + LED groups, thereby confirming that LED treatment does not adversely affect normal ocular function.

3. Fluorescein corneal staining (FCS) scores

On the 22nd day, FCS scores were significantly lower in the MGD + LED (p = 0.019) and MGD + R (p = 0.000) groups compared to the untreated MGD (Fig. 4B). While the MGD + L + H group demonstrated numerically lower scores compared to the MGD + H group, this difference was not statistically significant (p = 0.095). The study revealed no statistically significant differences between the CON and the CON + LED groups, thereby substantiating the non-damaging nature of LED therapy in healthy eyes.

Fig. 4

Ocular sign analysis on the MGD induced eyes. A Tear break up time (TBUT); B: Fluorescein corneal staining score (FCS); C:Eyelid swelling scores; D: Eyelid telangiectasia scores. The statistical analysis was performed using ANOVA for each week. # p<0.05, ## p<0.01, ### p<0.001 compared to MGD group. CON: control; CON+LED: control with LED treatment; MGD: meinomian gland dysfunction; MGD+LED: meibomian gland dysfunction with LED treatment; MGD+HA: meibomian gland dysfunction with hyaluronic acid; MGD+R: meibomian gland dysfunction with cyclosporine treatment; MGD+H: meinomian gland dysfunction with thermal treatment; MGD+L+H: meibomian gland dysfunction with LED and thermal treatment

4. Eyelid swelling scores

A significant reduction in eyelid swelling scores was observed in all treatment groups, with the exception of MGD + HA (p = 0.854), in comparison to the untreated MGD group (p < 0.05). As demonstrated in Fig. 4 C, the MGD + LED group displayed a significant decrease in swelling on days 15 and 22 (p = 0.019 and p = 0.002, respectively). In contrast, the MGD + L + H group exhibited the most substantial improvement on day 22 (p = 0.000).

On the 22nd day, no statistically significant differences were observed among the MGD + LED, MGD + R, MGD + H, and MGD + L + H groups, indicating comparable therapeutic efficacy. Moreover, the study revealed that there were no statistically significant differences between the CON and CON + LED groups, thereby supporting the hypothesis that LED treatment is ocularly safe in normal eyes.

5. Eyelid telangiectasia scores

As demonstrated in Fig. 4D, a general trend toward reduced eyelid telangiectasia scores over time was exhibited by all treatment groups. However, statistically significant reductions were only observed in the MGD + R (p = 0.000) and MGD + L + H (p = 0.014) groups compared to the MGD group.

The MGD + LED group exhibited no statistically significant difference from the MGD group (p = 0.556). A downward trend was observed in the MGD + H group beginning on day 15 (p = 0.518), but statistical significance was only achieved when combined with LED therapy (MGD + L + H; p = 0.021). No substantial alterations were observed between the CON and CON + LED groups.

6. Immunohistochemistry (IHC) findings

IHC analysis was performed using DAB staining to assess the relative expression of inflammatory cytokines IL-1β, IL-6, and TNF-α in the ocular tissues. DAB oxidation was visualized as brown coloration, indicative of positive expression of target cytokines (Figs. 5 and 6).

Fig. 5

Immunohistochemistry results of the LED phototherapy device on the ocular tissues. A: Cornea. B: Conjunctiva. CON: control; CON+LED: control with LED treatment; MGD: meibomian gland dysfunction; MGD+LED: meibomian gland dysfunction with LED treatment; MGD+HA: meibomian gland dysfunction with hyaluronic acid; MGD+R: meibomian gland dysfunction with cyclosporine treatment; MGD+H : meibomian gland dysfunction with thermal treatment; MGD+L+H: meibomian gland dysfunction with LED and thermal treatment

Fig. 6

Immunohistochemistry results of the LED phototherapy device on the ocular tissues. A: Meibomian gland. B: Retina. CON: control; CON+LED: control with LED treatment; MGD: meibomian gland dysfunction; MGD+LED: meibomian gland dysfunction with LED treatment; MGD+HA: meibomian gland dysfunctionwith hyaluronic acid; MGD+R: meibomian gland dysfunction with cyclosporine treatment; MGD+H: meibomian gland dysfunction with thermal treatment; MGD+L+H: meibomian gland dysfunction with LED and thermal treatment

In corneal tissue, both the CON and CON + LED groups exhibited low expression levels of all three cytokines, consistent with a healthy, non-inflamed ocular surface. In contrast, the MGD group exhibited significant DAB staining in the corneal epithelium and basement membrane regions, suggesting pronounced inflammatory responses (Fig. 5 A).

Treatment of the groups resulted in reduced cytokine expression in comparison to the untreated MGD group. The MGD + LED, MGD + R, and MGD + L + H groups demonstrated the most substantial decreases across all three cytokines. The investigation revealed that hyaluronic acid treatment (MGD + HA) led to a significant decrease in the expression of IL-1β and TNF-α, while cyclosporine (MGD + R) demonstrated a more pronounced ability to suppress IL-1β and IL-6. The thermal therapy groups (MGD + H and MGD + L + H) exhibited mild DAB staining in the corneal epithelium but demonstrated overall decreased cytokine expression, with the combined treatment group (MGD + L + H) showing particularly notable reductions.

A similar pattern was observed in conjunctival tissue. The CON and CON + LED groups exhibited minimal cytokine expression, while the MGD group demonstrated strong staining for IL-1β, IL-6, and TNF-α throughout the conjunctival epithelial and basement membrane layers, thereby confirming successful induction of inflammation (Fig. 5B). Among the treatment groups, eyes that received LED therapy (MGD + LED) exhibited a general reduction in the expression of all cytokines when compared to the MGD group. Treatment with hyaluronic acid led to a significant decrease in the expression of pro-inflammatory cytokines, such as IL-1β and TNF-α. This treatment also resulted in a noticeable reduction in the intensity of conjunctival epithelium and basal layer staining. In contrast, cyclosporine treatment demonstrated a more pronounced ability to suppress the expression of IL-1β and IL-6. Thermal therapy, particularly in combination with LED irradiation, resulted in a significant reduction of all three cytokines, despite the presence of some residual staining in the superficial epithelial region. The findings of this study suggest that LED therapy, when used in isolation or in combination with thermal treatment, exerts a significant anti-inflammatory effect on the ocular surface, comparable to the effects of conventional pharmaceutical interventions.

In the meibomian glands, the expression of inflammatory cytokines was minimal in the CON and CON + LED groups, with only faint DAB staining observed in the lobular and ductal regions (Fig. 6 A). Conversely, the MGD group demonstrated robust and pervasive staining of IL-1β, IL-6, and TNF-α across the acini and periductal regions, thereby substantiating the efficacy of the inflammation induction process. Among the treatment groups, MGD + LED demonstrated a reduction in the expression of all cytokines, while MGD + HA significantly decreased IL-1β and TNF-α levels. MGD + R exhibited superior suppression of IL-1β and IL-6, while MGD + H demonstrated more pronounced efficacy in reducing IL-6 and TNF-α. The combined treatment group (MGD + L + H) demonstrated the most significant reductions in all cytokines, particularly in the ductal epithelium.

Cytokine expression in retinal tissues remained minimal in the CON and CON + LED groups, with weak staining observed in the inner nuclear layer (INL), outer nuclear layer (ONL), and retinal pigment epithelium (RPE) (Fig. 6B). However, the MGD group exhibited augmented expression of IL-1β, IL-6, and TNF-α across multiple layers, suggesting that MGD-associated inflammation propagated to the posterior segment. Following treatment, MGD + LED demonstrated mild attenuation of all cytokines, especially in the ONL and RPE, while MGD + HA reduced IL-1β and TNF-α expression. MGD + R exhibited a more pronounced suppression of IL-1β and IL-6, while MGD + H demonstrated a comprehensive suppression of IL-6 and TNF-α across the retinal layers. The combined treatment group (MGD + L + H) exhibited a consistent decrease in cytokine expression, as evidenced by reduced staining in both the inner nuclear layer (INL) and the retinal pigment epithelium (RPE). These findings suggest that PBM therapy, particularly when combined with thermal treatment, effectively attenuates MGD-associated inflammation in both anterior and posterior ocular tissues.

7. H-score quantification of inflammatory cytokines

Quantitative analysis, employing pixelwise H-scores as a metric, revealed a marked upregulation of proinflammatory cytokines, including IL-1β, IL-6, and TNF-α, in the cornea, conjunctiva, meibomian gland, and retina of the MGD group when compared to the control group (Fig. 7). Notably, all treatment groups exhibited a general trend of decreased cytokine expression. In the cornea, significant reductions in IL-1β levels were observed in the MGD + LED, MGD + HA, MGD + R, and MGD + H groups. Concurrently, notable suppression of IL-6 and TNF-α was evident in the MGD + R and MGD + L + H groups, respectively. In conjunctival tissue, the levels of IL-1β were significantly decreased in the MGD + LED, MGD + R, MGD + H, and MGD + L + H groups, and IL-6 expression was significantly reduced in the MGD + LED and MGD + R groups. TNF-α demonstrated a downward trend across treatments, with statistical significance observed exclusively in the MGD + R group. In the meibomian glands, where all cytokines exhibited marked elevation in the MGD group, IL-1β demonstrated significant reduction in the MGD + H and MGD + L + H groups, while IL-6 exhibited significant suppression in all treatment groups. TNF-α demonstrated a downward trend across treatments, though this variation was not statistically significant. In retinal tissue, the expression of cytokines was found to be mildly elevated in the MGD group in comparison to the control group. While the overall changes were less pronounced, it was observed that all treatments resulted in a general reduction in cytokine expression. A statistically significant decrease in IL-6 was noted in the MGD + LED group. These findings indicate that pharmacological and non-pharmacological interventions, particularly combined LED and thermal therapy, effectively attenuate MGD-induced inflammation across ocular tissues.

Fig. 7

Immunohistochemistry results with H-scores. A. Cornea; B: Conjuctiva; C: Meibomian gland; D:Retina. *p<0.05, **p<0.01 compared to CON; # p<0.05, ## p<0.01, ### p<0.001 compared to MGD group. CON: control; CON+LED: control with LED treatment; MGD: meibomian gland dysfunction; MGD+LED: meibomian gland dysfunction; MGD+LED: meibomian gland dysfunction with LED treatment; MGD+HA: meibomian gland dysfunction with hyaluronic acid; MGD+R: meibomian gland dysfunction with cyclosporine treatment; MGD+H: meibomian gland dysfunction with thermal treatment; MGD+L+H: meibomian gland dysfunction with LED and thermal treatment

8. TUNEL staining and retinal safety

TUNEL staining was conducted to evaluate retinal cell apoptosis. No significant TUNEL-positive signals were detected in the CON + LED, MGD + LED, or MGD + L + H groups, indicating minimal or no DNA fragmentation. These results indicate that LED irradiation, whether utilized as a standalone modality or in conjunction with thermal therapy, does not result in retinal damage or apoptosis. This observation supports the ocular safety of this therapeutic approach (Fig. 8).

Fig. 8

TUNEL staining results on experiment groups with LED treatment. No significant TUNEL positive signals were detected in all the LED groups. CON+LED: control with LED treatment; MGD+LED: meibomian gland dysfunction with LED treatment;MGD+L+H: meibomian gland dysfunction