Optical density changes of subretinal hyperreflective material in age-related macular degeneration after switching therapy from ranibizumab to aflibercept.

BACKGROUND: Subretinal hyperreflective material (SHRM) is a hyperrefiective material seen on optical coherence tomography (OCT) and located under the retina and above the retinal pigment epithelium. This study aims to examine the effect of SHRM on the functional prognosis of age-related macular degeneration (AMD) patients who switched from intravitreal ranibizumab to intravitreal aflibercept treatment.
MATERIALS AND METHODS: This is a retrospective, nonrandomized clinical study. AMD patients meeting the switching criteria underwent a complete ophthalmic examination, including spectral-domain OCT and fundus fluorescein angiography. The best-corrected visual acuity and OCT parameters were measured at the switch and 3, 6, 12, and 24 months after. SHRM(+/-), maximum SHRM thickness, and subjective and objective reflectivity stages of SHRM (grades 1-3) were evaluated.
RESULTS: SHRM was observed in 24/48 (50.0%) of eyes at the time of the switch. The differences in maximum SHRM thicknesses were not statistically significant. SHRM's mean subjective reflectivity stages at the switch and subsequent examinations were 2.37, 2.75, 2.75, 2.74, and 2.81; SHRM's objective reflectivity staging also confirmed them. Functional changes after the switch showed a significant VA loss in the SHRM(+) group and significant gain in the SHRM(-) group.
CONCLUSION: This study showed that the presence of SHRM and higher optical reflectivity at the switch from ranibizumab to aflibercept caused a poor prognosis after the switch. On the other hand, SHRM(-) patients achieved good functional results after the switch. Copyright:

Introduction

Subretinal hyperreflective material (SHRM) is a complex material consisting of fluid, fibrin, hemorrhage, fibrosis, and choroidal neovascular membrane (CNVM) in varying proportions.[12] Previous studies reported a decrease in the fluid and fibrin contents of treatment-naïve neovascular age-related macular degeneration (AMD) patients receiving anti-vascular endothelial growth factor (anti-VEGF) treatment.[13] Besides, some patients were observed to have fibrosis in the follow-ups, and their visual acuity (VA) was lower.[45]

This study aims to evaluate the effect of SHRM on functional prognosis in AMD patients after a switch from intravitreal ranibizumab (IVR) to intravitreal aflibercept (IVA) treatment.

Materials and Methods

This is a retrospective, nonrandomized, clinical study. The clinical examination and optical coherence tomography (OCT) parameters of AMD patients who received IVA treatment after getting an inadequate response from IVR at Istanbul Training and Research Hospital between January 2016 and 2021 were retrospectively analyzed from patient records. The study was approved by Istanbul Training and Research Hospital's Clinical Trials Ethics Committee.

Treatment-naïve neovascular AMD patients received three monthly loading doses of IVR, and PRN protocol was followed. The following findings observed in monthly follow-ups were accepted as reinjection criteria, and additional IVR injection was performed: intraretinal fluid/cyst, subretinal fluid in OCT, leakage in fundus fluorescein angiography (FFA), new developing macular hemorrhage on fundus examination, and 1 line or more loss in best-corrected VA (BCVA) compared to the previous control.[67]

The following findings, observed after at least five IVR doses, were defined as inadequate IVR response: intraretinal fluid/cyst, subretinal fluid in OCT, leakage in FFA, new developing macular hemorrhage on fundus examination, and 1 line or more loss in BCVA compared to the previous control. IVR treatment was replaced by IVA in patients who gave an inadequate response. There was a 1-month interval between the last dose of IVR and the first dose of IVA.

After the switch, patients received three monthly loading doses of IVA, and PRN protocol was followed.[8] In the monthly follow-up, an additional IVA was performed on the patients who met the reinjection criterion.

Inclusion criteria

Patients over 50 years old, having CNVM secondary to AMD proven by OCT and FFA, having received IVR at least five times, and followed at least 24 months after the switch, were included.

Exclusion criteria

Presence of other diseases that may cause CNVM

Having diseases affecting the macula, such as diabetic retinopathy, epiretinal membrane, and vitreomacular traction

Undergoing vitreoretinal surgery

Undergoing intraocular surgery that may affect VA after the switch.

BCVA was evaluated monthly using a decimal visual chart, biomicroscopic funduscopy, and OCT. FFA was performed in the initial diagnosis of AMD and repeated, if necessary, to differentiate polyps before considering a switch.

BCVA and OCT were evaluated at the switch and 3, 6, 12, and 24 months after. Decimal VA was converted to logMAR VA. Regarding BCVA, 0.2 or more difference in logMAR was taken as clinically significant. A logMAR increase or decrease of <0.2 was accepted as clinically stable VA.

Measurements were made using spectral-domain OCT (RTVue 100-2; Optovue, Fremont, CA). Measurements with a reliability index of 60 and above were included.

The macula was imaged in a 5 mm × 5 mm area by OCT (MM5 mode). Maximum thickness and reflectivity stages of SHRM(+/−) were determined. OCT findings were evaluated by two independent graders (A.F. and H.G.). The Spearman correlation analysis was used to reveal the relationship between the two graders. The correlation coefficient was 0.986 (P = 0.000). Therefore, the average maximum SHRM thickness obtained from the two graders was used for analyses. The correlation coefficient was 1.000 (P = 0.000) in the subjective evaluation of reflectivity stages.

The hyperreflective material located under the retina above retinal pigment epithelium (RPE) was defined as SHRM. In cases where SHRM and RPE can be distinguished, maximum SHRM thickness is taken as the distance between the inner edge of RPE and the inner edge of SHRM.[14] On the other hand, if SHRM and RPE cannot be distinguished, maximum SHRM thickness was taken as the distance between the inner edge of the Bruch membrane and the inner edge of SHRM [Figure 1].[14]

Figure 1

OCT scans showing maximum SHRM thickness (a and b) SHRM and RPE can be distinguished (169 μm/140 μm) (c) SHRM and RPE cannot be distinguished (222 μm). SHRM: Subretinal hyperreflective material; RPE: Retinal pigment epithelium; OCT: Optical coherence tomography

The subjective reflectivity of SHRM was staged regarding RPE and the outer plexiform layer (OPL).[4] The stages are as follows: Grade 1 – the reflectivities of SHRM and OPL are similar, Grade 2 – the reflectivity of SHRM is between the RPE's and OPL's reflectivities, and Grade 3 – the reflectivities of SHRM and RPE are similar [Figure 2].

Figure 2

Subjective reflectivity staging of SHRM; (a) Grade 1: The reflectivities of SHRM and OPL are similar; (b) Grade 2: The reflectivity of SHRM is between the RPE's and OPL's reflectivities; (c) Grade 3: The reflectivities of SHRM and RPE are similar (SHRM: Subretinal hyperreflective material; RPE: Retinal pigment epithelium; OPL: Outer plexiform layer)

The objective reflectivity of SHRM was also evaluated. Three sections of 20 × 20 pixels were taken from RPE, OPL, and SHRM layers using the Adobe Photoshop 2020 program. The sections of 20 × 20 pixels consist of 400 frames, each with a pixel value between 0 and 255 representing the brightness. Sections’ mean pixel values were calculated in MATLAB Online R2020b program. A layer's mean pixel value is the average of the three sections cut from the layer. The images of 20 × 20 pixel sections cut from the RPE layer are shown in Figure 3. In this way, the pixel intervals and mean pixel values of SHRM reflectivity, which has been subjectively evaluated in OCT, were determined objectively.

Figure 3

Objective evaluation of the SHRM using the MATLAB Online R2020b program. (a) First section of retinal pigment epithelial layer. Mean pixels of 400 frames: 165.3 pixels; (b) second section of retinal pigment epithelial layer. Mean pixels of 400 frames: 174.8 pixels; (c) third section of retinal pigment epithelial layer. Mean pixels of 400 frames: 164.2 pixels. Mean pixels of retinal pigment epithelial layer: 168.1 pixels

Statistical analysis

Statistical analyses were performed using Statistical Package for the Social Sciences (SPSS, version 22.0). Wilcoxon test was used to compare maximum SHRM thickness at the time of the switch and after the switch. Mann–Whitney U-test was used to compare the groups in terms of VA changes after 24 months. Kaplan–Meier method was used in survival analysis. Statistical significance level was taken as P < 0.05.

Results

The demographic characteristics of the patients, follow-up information, and the number of injections are shown in Table 1.

Table 1

Demographic characteristics of the patients, follow-up information, and the number of injections

Number of patients/eyes (n)	47/48
Age (years) (n)	74.1±7.2 (59-89) (47)
Number of females/males (n)	22/25
Mean follow-up time before switch (months)	18.8±8.2 (5-49)
Average number of intravitreal ranibizumab before switch	9.3±4.0 (5-27)
Average number of intravitreal aflibercept in 24 months after switch	8.9±3.0 (3-17)

SHRM was present in 24 (50.0%) eyes at the time of the switch; it was resorbed after 12 months in 1 eye and resorbed after 24 months in another eye. The maximum SHRM thicknesses at the time of the switch and 3, 6, 12, and 24 months after the switch are shown in Table 2.

Table 2

Maximum subretinal hyperreflective material thicknesses, subretinal hyperreflective material reflectivities, and pixel values of subretinal hyperreflective material reflectivities at the time of the switch, and 3, 6, 12, and 24 months after the switch

Parameters	Mean±SD (minimum-maximum)
	Switch time	3rd month after switch	6th month after switch	12th month after switch	24th month after switch
Maximum SHRM thickness (µm)	181.2±58.8 (95-305)	167.2±63.9 (87-317)	166.4±58.9 (78-271)	167.5±68.1 (0-259)	163.2±79.0 (0-292)
Pixel values of SHRM reflectivities (pixels)	Grade 1 (n=4) 94.3±6.2 (85.1-98.4) Grade 2 (n=7) 125.1±6.8 (116.1-133.7) Grade 3 (n=13) 168.1±8.7 (153.2-186.2)	Grade 2 (n=6) 123.9±9.3 (118.4-137.3) Grade 3 (n=18) 166.8±9.4 (151.7-189.0)	Grade 2 (n=6) 124.6±11.0 (116.8-141.3) Grade 3 (n=18) 168.4±8.0 (152.2-188.8)	Grade 2 (n=6) 126.4±5.6 (116.5-134.3) Grade 3 (n=17) 168.8±8.6 (155.2-187.7)	Grade 2 (n=4) 126.7±11.7 (117.9-143.0) Grade 3 (n=18) 167.3±8.8 (155.3-186.1)
SHRM reflectivities	Grade 1 (n=4) Grade 2 (n=7) Grade 3 (n=13) Mean grade=2.37 (n=24)	Grade 2 (n=6) Grade 3 (n=18) Mean grade=2.75 (n=24)	Grade 2 (n=6) Grade 3 (n=18) Mean grade=2.75 (n=24)	Grade 2 (n=6) Grade 3 (n=17) Mean grade=2.74 (n=23)	Grade 2 (n=4) Grade 3 (n=18) Mean grade=2.81 (n=22)

SHRM: Subretinal hyperreflective material, SD: Standard deviation, n: Number of eyes

The reduction in maximum SHRM thickness was not statistically significant (3rd month, P = 0.399; 6th month, P = 0.637; 12th month, P = 0.494; and 24th month, P = 0.502).

Regression analysis showed no significant reduction in maximum SHRM thickness in the follow-up (P = 0.097, r2 = 0.655). 77.7% of maximum SHRM thickness reduction occurred in the 3rd month.

Sections of 20 × 20 pixel size were examined for the objective reflectivity staging of SHRM. The mean pixel values were as follows: OPL, 87.0 ± 3.0 (81.5–96.7); RPE, 162.7 ± 5.7 (151.3–175.5); and SHRM, 154.9 ± 23.0 (85.1–189.0). The objective reflectivity of SHRM was staged regarding RPE and OPL points. Accordingly, SHRM pixel values under 100.0 are Grade 1, 100.0–149.9 are Grade 2, and over 150.0 are Grade 3.

The SHRM reflectivities (n = 24) at the time of the switch and 3, 6, 12, and 24 months after the switch are shown in Table 2. Regression analysis showed no significant increase in reflectivity stages of SHRM at follow-up (P = 0.124, r2 = 0.599). 86.3% of the increase in SHRM reflectivity stages occurred in the 3rd month.

The means of BCVA in the SHRM(+) and SHRM(−) groups at the time of the switch and after 24 months and the comparison of the groups are shown in Table 3. The mean VA loss was 0.38 ± 0.40 logMAR in the SHRM(+) group, whereas the mean VA gain was 0.21 ± 0.33 logMAR in the SHRM(−) group (P = 0.000).

Table 3

The means of best-corrected visual acuity in subretinal hyperreflective material (+) and subretinal hyperreflective material (−) groups at the time of the switch and after 24 months and the comparison of the groups

Groups	Mean±SD (n=24)
	Switch time, visual acuity (logMAR)	24th month after switch, visual acuity (logMAR)	Visual acuity change (logMAR)
SHRM(+) group	1.04±0.38	1.43±0.61	+0.38±0.40
SHRM(−) group	0.92±0.30	0.71±0.40	−0.21±0.33

P=0.000; Mann-Whitney U-test: Comparison of SHRM(+) group and SHRM(−) group in terms of visual acuity change. SHRM: Subretinal hyperreflective material, SD: Standard deviation, n: Number of eyes, logMAR: Logarithm of the minimum angle of resolution

In the SHRM(+) group, VA increased in 2 eyes, remained stable in 4 eyes, and decreased in 18 eyes after 24 months. Regarding the SHRM(−) group, VA increased in 14 eyes, remained stable in 6 eyes, and decreased in 4 eyes after 24 months.

The effect of SHRM on functional prognosis was evaluated with survival analysis and multiple regression analysis. Regarding survival analysis, VA increased in 2 out of 24 eyes in the SHRM(+) group. On the other hand, VA increased in 14 of 24 eyes in the SHRM(−) group (P = 0.000). A statistically significant difference was found in multiple regression analysis (P = 0.000, r2 = 0.404).

The correlation between SHRM reflectivity stages and VA at the time of the switch could not be evaluated in the SHRM(+) group due to the low number of patients.

Discussion

In previous studies involving anti-VEGF treatment, SHRM has been observed to resorb in some patients, and the most significant reduction was observed in the 1st month after the treatment.[19] In particular, this reduction in SHRM was attributed to a decrease in its fluid or fibrin content.[19] In the study of Shah et al.,[3] SHRM, of which fibrin is the main component, regressed with anti-VEGF treatment, and VA increased with the regression.

In their study, Casalino et al.[9] reported that the increased reflectivity of SHRM observed in spectral-domain OCT indicated fibrosis. The study by Kumar et al.[4] found fibrosis in the patients with more hyperreflectivity of SHRM in the spectral domain OCT, and VA was lower in these patients. Previous studies also emphasized that fibrosis and scarring are the factors that negatively affect VA.[1101112]

In their study involving the treatment-naïve neovascular AMD patients, Roberts et al.[5] reported using polarization-sensitive OCT that SHRM content changed after three doses of anti-VEGF treatment from a vascular form to a fibrotic form in some patients, and their VA was lower. Diagnosing fibrosis in SHRM by using spectral-domain OCT is more difficult. The patients in the current study had previously received IVR. Therefore, the fluid and fibrin content of SHRM was likely to decrease, and fibrotic components’ rate to increase. In addition, studies conducted by Willoughby et al.[1] and Roberts et al.[5] have indicated that low reduction in SHRM thickness after anti-VEGF is associated with the development of fibrosis. In the current study, the absence of a significant change in maximum SHRM thickness after the switch suggests that the SHRM had a high fibrotic component and a low fluid and fibrin component at the time of the switch.

A further reduction in maximum SHRM thickness occurred at the 3rd month, showing that the fluid and fibrin, found in small amounts in the SHRM at the switch, were resorbed. SHRM thickness varied between 167.2 and 163.2 μm between the 3rd and 24th months, indicating subretinal fibrosis development from 3 months after the switch. Furthermore, the mean of SHRM's reflectivity stages was 2.37 at the switch, indicating a high fibrotic ratio in SHRM. The mean of SHRM's reflectivity stages increased from 2.37 to 2.75 in 3 months, which means that the fluid and fibrin, forming a small part of the SHRM content at the time of the switch, were resorbed. The mean of SHRM's reflectivity stages in the 3rd and 24th months was 2.74 and 2.81, indicating the development of subretinal fibrosis after the 3rd month.

Conclusion

The present study demonstrated that the presence of SHRM and the intensity of optical reflectivity at the time of the switch negatively affect the prognosis. Since the SHRM present at the switch had a high fibrotic component and low fluid and fibrin components, the prognosis may be poor in SHRM(+) patients, despite the switch. However, the SHRM(−) group gained significant visual recovery after 24 months.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.