Impact of COVID-19-related lockdown on retinal disorders treated with intravitreal injections.

PURPOSE: To study functional changes in vision and morphological changes on optical coherence tomography (OCT) scans in patients where intravitreal therapy (IVT) with injections has been delayed due to COVID-19 pandemic lockdown.
METHODS: Retrospective cross-sectional study included 77 eyes with the diagnosis of exudative age-related macular degeneration (eAMD), diabetic macular edema (DME), and retinal vein occlusion (RVO), regularly receiving IVT with anti-vascular endothelial growth factor (Anti-VEGF) injections or dexamethasone implant and were not able to receive the injections as intended due to lockdown. Best-corrected visual acuity (BCVA), central foveal thickness (CFT), and qualitative morphological features on OCT were analyzed pre- and post-therapy break off.
RESULTS: The mean duration of IVT break-off was 57 days. Mean BCVA logarithm of minimum angle of resolution (logMAR) worsened from 0.33 (20/42) to 0.35 (40/44) (P = 0.02) and mean CFT increased from 297.90 μm to 402.16 μm (P < 0.01), from pretherapy break-off to return visit. Worsening of the visual acuity was seen across all the three disease cohorts, significantly more in the eyes with eAMD.
CONCLUSION: Marginal decline in the visual acuity and worsening of the OCT features were observed as a result of delay in the IVT injections of DME, eAMD and RVO patients. Copyright:

Introduction

Intravitreal therapy (IVT) anti-vascular endothelial growth factor (VEGF) injections have been the mainstay of treatment in various retinal disorders including AMD, retinal vein occlusion (RVO), and diabetic macular edema (DME).[12345] Intravitreal steroid implants are also an established treatment modality for RVO and DME.[67] Both the modalities improve visual acuity by reducing the amount of retinal fluid, however, IVT requires regular and frequent hospital visit depending upon the duration of action of the intravitreal drug. Studies have repeatedly shown the need of frequent injections for the preservation of vision.[89] It is well-established that patients who have poor compliance and hence are lost to follow-up have worse visual acuity (VA) outcome.[10] Prime reasons attributed to noncompliance were lack of financial funding and psychological burden and comorbid systemic diseases[111213]

COVID-19 pandemic is a rare situation that has resulted in countries enduring strict lockdowns to control the spread that has adversely affected the routine hospital visits of patients of other diseases. In the current study, we aim to study the morphological changes and visual outcomes after a gap in clinical care due to such unavoidable circumstances. We aimed to evaluate the functional and anatomical outcomes of eyes ho had a break-off in IVT as a result of the pandemic-induced travel restrictions.

Methods

The study was reviewed and approved by the Medical Research and Ethics Committee at Grewal Eye Institute, Chandigarh. The Institutional Review Board approved the study and it adhered to the tenets of the Declaration of Helsinki.

Study population

Retrospective cohort study included patients who were receiving intravitreal injections for various retinal disorders on regular basis and could not report to the hospital on the intended date of follow-up due to lockdown. Patients with the initial diagnosis of exudative age-related macular degeneration (eAMD), DME and RVO who were receiving IVT with Anti-VEGF injections monthly/bimonthly and IVT with intravitreal dexamethasone implants three monthly were included. The patient cohort was identified from the electronic medical record (EMR) of the IVT and online appointment list for medical retina patients. At the end of every clinical consultation and last IVT, the next follow-up appointment was routinely scheduled and noted in EMR. Only those cases who had an IVT breakoff of minimum 14 days from the intended day of treatment were included.

Baseline characteristics, defined as the last visit before the patient had IVT break off, included age, gender, retinal diagnosis, and type of IVT injection were recorded. Visual acuity recorded at the last visit before IVT and return visit after therapy break off was noted. Visual acuity was recorded using Snellen VA and was converted to LogMAR for analysis. All patients included in the study had fundus photography, fundus fluorescein angiography and/or optical coherence tomography (OCT) Angiogram for confirmation of retinal diagnosis at baseline of treatment initiation. The intravitreal treatment protocol was loading monthly three doses for all Anti-VEGFs followed either by fixed interval regime or pro re nata (PRN) basis as decided by the treating vitreoretina consultant.

OCT scans were performed using spectral-domain OCT (Cirrus 5000)/PLEX® Elite 9000, Carl Zeiss Meditec, Inc., Dublin, CA, USA. OCT scan at baseline was analyzed by a single retinal specialist for the presence or absence of subretinal fluid (SRF)/intraretinal fluid (IRF)/pigment epithelial detachment (PED) and central foveal thickness (CFT). CFT was measured from the inner border of the retina to the inner border of the hyperreflective layer of RPE using calipers in machine software. OCT scans were reanalyzed at the return visit after breakoff of IVT. Patients were excluded if the OCT scans were not available or the signal strength was <5.

For statistical analysis, Microsoft Excel 2013 and SPSS software (Version 23, IBM, India) were used. The Snellen VA was converted into logMAR for the analysis. Descriptive statistics were applied to the various characteristics of patients. All categorical and continuous comparisons were performed using a Chi-square test and one-way analysis of variance, respectively. Paired t-test and Wilcoxon signed-rank tests were used to find the difference between means of pre and post visits parameters for parametric and nonparametric data, respectively. P <0.05 was considered statistically significant.

Results

Data for 79 eyes of 68 patients were collected. Two eyes were excluded from the study as one developed vitreous hemorrhage and the other did not have an OCT scan at the return visit. Baseline characteristics of 77 eyes of 66 patients were included for analysis [shown in Table 1]. 29 eyes (37.6%) had a diagnosis of eAMD, 33 eyes (42.8%) had DME, and 15 eyes (19.4%) had RVO.

Table 1

Baseline characteristics stratified by anti-vascular endothelial growth factor agent

n=77 eyes	Ranibizumab (36 eyes)	Bevacizumab (20 eyes)	Aflibercept (11 eyes)	Dexamethasone (10 eyes)	P
Gender
Male	21	11	6	4	0.46*
Female	9	5	5	5
Age (years) mean (SD)	71.24 (13.10)	67.43 (12.40)	65.81 (9.65)	66.33 (7.42)	0.49†
Mean number of injections before IVT break off (SD)	10.85 (12.42)	6.8 (3.90)	9.54 (5.38)	3.6 (3.38)	0.12†
Number of days of IVT breakoff mean (SD)	56.5 (19.08)	55.3 (21.40)	56.81 (19.06)	63.9 (16.62)	0.72†

*Chi-square test, † One-way ANOVA. SD: Standard deviation, IVT: Intravitreal therapy, ANOVA: Analysis of variance

Mean age was 68.71 years (standard deviation [SD] =12.07). 42 patients (63%) in the cohort were males and 24 (37%) were females. Mean number of intravitreal anti VEGF injections (Ranibizumab, Bevacizumab, Aflibercept) received before IVT break off were 10.85 , 9.54 and 6.8 respectively. The mean number of intravitreal dexamethasone implants given to the patients with DME and RVO before therapy break off were 3.6 (SD = 3.38). Mean duration of therapy break off (i.e. the time interval between the intended dates of follow-up visit to return visit date after lockdown) was 57 (17–89) days. Mean duration of therapy break off for patients receiving Ranibizumab, Bevacizumab, Aflibercept and dexamethasone implant was 56.5 days (SD = 19.08), 55.3 days (SD = 21.40), 56.81 days (SD = 19.06), and 63.9 days (SD = 16.62), respectively.

Changes in visual acuity

Overall, mean logMAR VA worsened from visit before IVT break off 0.33 (20/42) to 0.35 (20/44) at return visit (P = 0.02).

Mean logMAR VA worsened in eyes that received ranibizumab from visit before therapy break-off (0.32 or 20/41) to return visit (0.38 or 20/47) (P = 0.31). Mean logMAR VA worsened significantly in eyes that received bevacizumab from visit before therapy break-off (0.42 or 20/52) to return visit (0.46 or 20/57) (P = 0.04). Mean logMAR VA worsened in eyes that received aflibercept from visit before therapy break-off (0.30 or 20/39) to return visit (0.50 or 20/63) (P = 0.05). Mean logMAR VA worsened in eyes that received dexamethasone from visit before therapy break-off (0.31 or 20/40) to return visit (0.61 or 20/81) (P = 0.1) [Table 2].

Table 2

Optical coherence tomography characteristics stratified by intravitreal anti-vascular endothelial growth factor agent at the visit before intravitreal therapy break off and return visit

n=77 eyes	Ranibizumab (36 eyes)	Bevacizumab (20 eyes)	Aflibercept (11 eyes)	Dexamethasone (10 eyes)	P
Visit before IVT break off
CFT in mm (SD)	264.41 (75.99)	334.95 (106.18)	290.18 (144.27)	352.9 (102.24)	0.03*
Intraretinal fluid (%)	17 (47.2)	15 (75)	2 (18)	8 (80)
Subretinal fluid (%)	5 (13.88)	8 (40)	8 (72)	1 (10)
LogMAR VA	0.32 (20/41)	0.42 (20/52)	0.30 (20/39)	0.31 (20/40)	0.64*
Return visit
CFT in mm (SD)	361.52 (101.61)	405 (153.72)	419.54 (213.44)	523.7 (159.66)	0.02*
Intraretinal fluid (%)	18 (50)	15 (75)	2 (18)	8 (80)
Subretinal fluid (%)	9 (25)	5 (25)	11 (100)	1 (10)
LogMAR VA	0.38 (20/47)	0.46 (20/57)	0.50 (20/63)	0.61 (20/81)	0.86*

*One-way ANOVA. CFT: Central foveal thickness, SD: Standard deviation, LogMAR: Logarithm of minimum angle of resolution, VA: Visual acuity, IVT: Intravitreal therapy, ANOVA: Analysis of variance

Mean logMAR VA worsened in eyes with eAMD from visit before therapy break-off 0.33 (20/42) to return visit 0.43 (20/53) (P = 0.04). Mean logMAR VA was comparable in eyes with DME from visit before therapy break-off 0.32 (20/41) to return visit 0.34 (20/43) (P = 0.32). Mean logMAR VA worsened in eyes with RVO from visit before therapy break-off 0.41 (20/51) to return visit 0.50 (20/63) (P = 0.25) [Table 3].

Table 3

Optical coherence tomography characteristics stratified by disease group/cohort at the visit before intravitreal therapy break off and return visit

n=77 eyes	eAMD (29 eyes)	DME (33 eyes)	RVO (15 eyes)	P
Mean umber of injections before IVT break off	12.45	7.48	7.53
Number of days of IVT break off mean (SD)	57.83 (20.04)	56.58 (19.75)	57.93 (20.33)
Visit before IVT break off
CFT in mm (SD)	288.48 (126.46)	304.06 (94.50)	302.6 (95.57)	0.83*
Intraretinal fluid (%)	9 (31)	25 (75.75)	8 (53.3)
Subretinal fluid (%)	19 (65.5)	2 (6)	1 (6.6)
Visual acuity (logMAR)	0.33 (20/42)	0.32 (20/41)	0.41 (20/51)	0.58*
Return visit
CFT in mm (SD)	372.48 (179.81)	409.21 (143.55)	444.07 (117.34)	0.32*
Intraretinal fluid (%)	13 (44.8)	22 (66.6)	8 (53.3)
Subretinal fluid (%)	23 (79.3)	2 (6)	1 (6.6)
Visual acuity (logMAR)	0.43 (20/53)	0.34 (20/43)	0.50 (20/63)	0.33*

*One-way ANOVA. CFT: Central foveal thickness, LogMAR: Logarithm of the minimum angle of resolution, PED: Pigment epithelial detachment, SD: Standard deviation, IVT: Intravitreal therapy, eAMD: Exudative age-related macular degeneration, DME: Diabetic macular edema, RVO: Retinal vein occlusion

In the AMD cohort, patients who received ranibizumab, bevacizumab and aflibercept, the mean logMAR VA changed from 0.28 (20/38) to 0.4 (20/50); 0.45 (20/56) to 0.52 (20/66) and 0.3 (20/39) to 0.38 (20/47), respectively, from the visit before IVT break off to the return visit. In the DME cohort, patients who received ranibizumab, bevacizumab, and dexamethasone, the mean logMAR VA changed from 0.3 (20/39) to 0.29 (20/38); 0.37 (20/46) to 0.38 (20/47) and 0.28 (20/38) to 0.38 (20/47), respectively, from the visit before IVT break off to the return visit. In the RVO cohort, patients who received ranibizumab, bevacizumab, and dexamethasone, the mean logMAR VA changed from 0.41 (20/51) to 0.52 (20/66); 0.42 (20/52) to 0.5 (20/63) and 0.4 (20/50) to 0.4 (20/50), respectively, from the visit before IVT break off to the return visit [Table 4].

Table 4

Visual acuity changes stratified by disease group/cohort at the visit before intravitreal therapy break off and return visit

Cohort	Mean logMAR VA	Ranibizumab	Bevacizumab	Afibercept
AMD	VA pre	0.28 (20/38)	0.45 (20/56)	0.30 (20/39)
	VA post	0.40 (20/50)	0.52 (20/66)	0.38 (20/47)
DME	VA pre	0.30 (20/39)	0.37 (20/46)	0.28 (20/38)
	VA post	0.29 (20/38)	0.38 (20/47)	0.38 (20/47)
RVO	VA pre	0.41 (20/51)	0.42 (20/52)	0.40 (20/50)
	VA post	0.52 (20/66)	0.50 (20/63)	0.40 (20/50)

AMD: Age-related macular degeneration, DME: Diabetic macular edema, RVO: Retinal vein occlusion, LogMAR: Logarithm of the minimum angle of resolution, VA: Visual acuity

Anatomical changes

Mean CFT increased from the visit before IVT break off 297.90 μm (82–580) to 402.16 μm (217–885) at return visit (P < 0.01). IRF and SRF were increased in 57.14% and 36% of eyes.

Mean CFT increased significantly from 262.42 μm to 361.53 in eyes treated with ranibizumab (P < 0.01); from 334.95 μm to 405 in eyes treated with bevacizumab (P < 0.01); from 290.18 μm to 419.55 in eyes treated with aflibercept (P = 0.03) and from 352.9 μm to 523.7 in eyes treated with dexamethasone (P = 0.04) [Table 2].

Mean CFT increased significantly from 302.6 μm to 444.07 in eyes with RVO (P < 0.01); from 340.06 μm to 409.21 in eyes with DME (P < 0.01) and from 288.48 μm to 372.48 in eyes with eAMD (P < 0.01) [Table 3].

PED was present in 24 out of 33 eyes (72.7%) with AMD and the mean PED height has found to be increased in 13 eyes (54.16%) after therapy break off as compared to before therapy break off; PED height has decreased in 3 (12.5%) eyes and remain unchanged in 8 eyes (33.33%). Mean linear dimensions changed from 2526 um to 3472 um after therapy break off.

In the AMD cohort, patients who received ranibizumab, bevacizumab, and aflibercept, the mean CFT changed from 270.9–339.81 μm; 313.42–349.85 μm and 290.18–419.54 μm, respectively, from the visit before IVT break off to the return visit. In the DME cohort, patients who received ranibizumab, bevacizumab, and dexamethasone, the mean CFT changed from 256.35 to 351.35 μm; 353.12–401 μm and 356.37–540.37 μm respectively, from the visit before IVT break off to the return visit. In the RVO cohort, patients who received ranibizumab, bevacizumab, and dexamethasone, the mean CFT changed from 272.62 to 413 μm; 336–488.6 μm and 339–457 μm respectively, from the visit before IVT break off to the return visit.

We also observed that in patients who had a delay in IVT of <45 days, the mean change in CFT was 60 μm as compared to the other group of patients with mean delay of more than 45 days where mean change in CFT was 125 μm.

Discussion

In the current study, we report the short-term results on vision and OCT changes as a result of strict country-wide lockdown due to the COVID-19 pandemic. Intravitreal injections are mandatory for the treatment of macular edema due to neovascular AMD/RVO/DME. Thirty percent of the patients who were receiving intravitreal injections in the past and had delayed treatment due to travel restrictions due to COVID-19 created lockdown reported worse mean logMAR VA at the return visit. These chronic diseases usually need long-term and regular treatment. Most patients are maintained either on PRN basis/fixed treatment regimen or treat and extend to resolve fluid and attain good VA. However, such strict protocols are difficult to achieve in real-world resulting in under treatment leading to visual loss.[141516]

Our results showed that an average delay of 57 days in IVT injections could result in a significant accumulation of fluid and cause worsening of vision. Studies in the past have noted the frequency and the causes for the long-term follow-up loss in patients receiving IVT injections for RVO/AMD/DME.[1718] Another study on patients with DME by Matsunaga et al.[19] had reported significant worsening in VA and CFT after lost to follow-up of 6 months or more. We analyzed the short-term delay in the IVT and noticed similar results suggesting that treatment break can lead to worsening in VA. However, it was beyond the limit of this study to analyze whether such changes are reversible or not.

The study by Rebecca et al.[20] showed that the mean duration of loss to follow-up (LTFU) for 346 days could result in logMAR VA drop from 0.6 (20/79) to 1 (20/200). Worsening of VA was found to be statistically significant in our study but was comparatively much less than the worsening of vision observed by Soares et al. Possible explanation for this could be a longer duration of LTFU for patients in their study. They also reported significant VA decline that persisted on follow-ups despite normalization of retinal thickness after the restoration of anti-VEGF injections.

Our study suggested some variations in the outcomes depending on the anti-VEGF injections used, but further exploration with a larger number of patients and longer follow-up will be required. We also observed that the patients receiving bevacizumab had a significant decline in vision at the return visit as compared to ranibizumab/Aflibercept. It is not entirely clear why the bevacizumab group had a larger decline in the VA, but we hypothesize that the eyes receiving bevacizumab had the poorer vision at the last follow-up before therapy break off as compared to the other two anti-VEGF arms. Cochrane-based metanalysis review of patients with DME have shown moderate certainty evidence that Aflibercept confers some advantage over ranibizumab and bevacizumab; however, we did not find any such advantage.[21]

The worsening of VA and CFT was seen across all the three disease cohorts [Figures 1-3], however, VA worsening was significant in only eAMD group. We did not observe any correlation between changes in the VA and CFT. In general, CFT is thought to be correlated with visual deterioration,[10] however, studies have also demonstrated that OCT findings often do not correlate well with the VA.[2223]

Figure 1

(a) Color fundus photo of a case of exudative AMD at baseline visit. (b) Color fundus photo at the last visit before therapy break off. (c) Color fundus photo at recent visit after therapy break off. (d) optical coherence tomography scan shows subretinal hyporeflective CNVM membrane complex with associated subretinal fluid and intraretinal fluid. (e) optical coherence tomography scan at the last visit before therapy break off. No retinal fluid was seen. (f) optical coherence tomography scan at recent visit after therapy break off of 71 days shows recurrence of subretinal fluid/intraretinal fluid.

Figure 3

(a) Color fundus photograph demonstrating multiple superficial hemorrhage, dilated tortuous retinal vessels and dull reflex at fovea in a case of CRVO. (b) Color fundus photo after 5 months of Anti- vascular endothelial growth factor therapy shows significant reduction in the hemorrhages. (c) Color fundus photograph at return visit after therapy break off of 33 days shows reappearance of intraretinal hemorrhages much worse than the previous two visits. (d) Standard deviation - optical coherence tomography at baseline shows macular edema in a case of CRVO with central foveal thickness of 1090 μm. (e) Standard deviation - optical coherence tomography scan demonstrates significant resolution of macular edema with central foveal thickness of 216 μm. (f) Standard deviation - optical coherence tomography scan shows reappearance of macular edema and increase in central foveal thickness to 1048 μm. Also not the presence of disorganization of the retinal inner layers

(a) Color fundus photograph of a case of moderate NPDR with diabetic macular edema at the time of initial diagnosis. (b) Color fundus photo after 4 months of 4 weekly ranibizumab injection. Note the reduction in the retinal hemorrhage and hard exudates. (c) Color fundus photo after therapy break off of 82 days. (d) Standard deviation - optical coherence tomography scan at baseline shows serous detachment at the fovea, intraretinal fluid, retinal thickening, hard exudates. central foveal thickness recorded was 82 μm. (e) Standard deviation - optical coherence tomography scan demonstrating significant reduction in subretinal fluid/intraretinal fluid after intravitreal therapy injection. central foveal thickness recorded was 373 μm. (f) Standard deviation -optical coherence tomography scan shows reappearance of subretinal fluid fluid pocket and significant increase in the amount of intraretinal fluid as a result of delay in intravitreal therapy therapy.

Our study had many limitations. Our disease cohorts were not limited to one particular pathology, rather we included three disease cohorts. We only analyzed the short-term effects in a limited number of patients and the long-term impact of VA and OCT changes still seem uncertain and beyond the scope of this study. Studies with longer follow-up and the larger number of patients will be able to assess the impact and reversibility of vision and OCT changes in case of lost to follow-up.

Conclusion

Our study concluded that even though the worsening of VA was marginal, but the difference was found to be statistically significant. We concluded that even a short-term delay in IVT injections could result in the worsening of the OCT features as well as functional vision deterioration. Hence, the patients receiving IVT injections are encouraged to adhere to regular follow-up and the treatment regimen as advised. Future research needs to work for the development of longer-acting injections which will help to combat the treatment burden and need for frequent visits associated with the current anti-VEGF therapy.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.