KOL KNOCKOUT™ MEDICAL RETINA EDITION
Vying for Superior Patient Outcomes in Retinal Care

This supplement contains highlights from two live symposia tailored for comprehensive ophthalmologists and retinal specialists who manage patients with retinal diseases. Key opinion leaders offer their diagnostic and therapeutic approaches for real patient cases, ranging from common to rare retinal disorders, while providing the clinical rationale behind their complex decision-making processes.

— Sunil K. Srivastava, MD

ROUND 1 | CASE 1: SELECTING INITIAL THERAPY IN A PATIENT WHO DEVELOPS WET AMD 

Dr. Srivastava: For the following four cases, I’m joined by Sruthi Arepalli, MD; Durga Borkar, MD, MMCi; and Ananth Sastry, MD. Our first patient is a 65-year-old woman who presented with blurred vision for 1 month in her right eye, with no prior history of eye issues. She had an aunt with age-related macular degeneration (AMD). She had mild nuclear sclerosis OU, and a VA of 20/50 OD and 20/30 OS. Her OCT scan showed a subretinal hemorrhage in the macula of her right eye and retinal changes associated with AMD in her left eye (Figure 1). Dr. Arepalli, do you need any other information to develop a management plan?

Figure 1. OCT imaging of right and left eye at presentation.

Sruthi Arepalli, MD: At my satellite offices, I typically don’t have the greatest fluorescein angiography (FA) or OCT angiography equipment, so I make most clinical decisions based on the OCT scan. I think we have plenty of good baseline information for this patient. When patients have a subretinal hemorrhage (indicating active disease) and their VA is 20/50 or worse, I usually start treatment with intravitreal aflibercept 2 mg injections. 

In general, I prefer off-label bevacizumab for the first three injections as first-line treatment for most patients with neovascular AMD (nAMD). Then, I usually have good success with medical insurances approving a medication switch to aflibercept 2 mg. However, in locations where follow-up may be difficult, I select a stronger drug initially. Being at a satellite clinic, I am most worried about patients being lost to follow-up. 

Dr. Srivastava: So, you choose aflibercept 2 mg when there are no barriers to treatment, such as insurance coverage. When would you bring this patient back? 

Dr. Arepalli: I generally start treatment intervals for intravitreal injections at every 4 weeks and then extend the interval if they’re responding favorably as evidenced by resolution of subretinal fluid (SRF). For this patient, I would tolerate some SRF if her hemorrhage is resolving.

Ananth Sastry, MD: Without any barriers to treatment, I also like to start with aflibercept 2 mg. If I observe any signs of pachychoroid spectrum disease, I may first reach for faricimab, which targets angiopoiten-2 and VEGF, anecdotally. Additionally, a recent paper by Tanaka et al showed that most treatment-naïve patients with pachychoroid spectrum disease receiving faricimab had resolution of SRF.¹ Otherwise, I reach for aflibercept 2 mg as my first treatment of choice.

Durga Borkar, MD, MMCi: In a world with no barriers, I would choose aflibercept 8 mg. I think for patients with larger pigment epithelial detachments (PEDs) I might consider treating with faricimab because there are some data to suggest that it can reduce larger PEDs that carry the risk of a tear.^2,3 Here, I’m not as worried about a tear and would also bring this patient back in 4 weeks. I also think that a FA is not needed in this case; her clinical presentation and OCT scans are sufficient. More specifically, this patient’s left eye clearly has OCT features of nAMD, which convinces me of her diagnosis. 

I typically load patients with nAMD who present with a subretinal hemorrhage every 4 weeks. For patients with less SRF, a smaller PED, and no submacular hemorrhage, I would consider extending the interval quicker. I think retina specialists may be somewhat “too hard and fast” on definitively treating with three or four loading doses, although I would in this case.

Dr. Srivastava: Does the presence of a PED influence your choice of drug for this patient? 

Dr. Borkar: The elevation of this PED is not high; I am less worried about a retinal pigment epithelium (RPE) rip. Therefore, I wouldn’t choose a less effective agent here. 

Dr. Sastry: I agree, RPE rips are a relatively rare occurrence from my clinical experience and respond well to treatment. However, I definitely warn my patients about it, especially if I see large PEDs. In this case, I also would not change my choice of medication because my goal is for both the SRF and hemorrhage to be resolved.

Dr. Srivastava: We all agree to start this patient on bevacizumab because of insurance requirements. Remember the importance of recommending AREDS supplementation to patients with moderate AMD or worse.^4,5 

At her 1-month follow-up after one injection of bevacizumab, this patient had a VA of 20/20 with some resolution of fluid. After her second injection of bevacizumab, she had subjective worsening of vision, and her VA was worse at 20/50. Her OCT scan showed a dark and rolled edge of RPE, indicating an RPE rip (Figure 2). We did warn this patient of the potential for an RPE tear, but her response was, “I was better last month until you gave me the injection.” So, would you change your treatment for this patient because of her RPE rip?

Figure 2. OCT imaging of right eye s/p 1 month after off-label bevacizumab x 2.

Dr. Borkar: I would recommend continuing treatment for this patient because her disease is still active. Certainly, it is essential to counsel patients. Managing chronic diseases can be challenging. Without treatment, this patient would likely have a VA worse than 20/50 due to worsening of the hemorrhage and SRF as well as the potential for fibrosis. Although she didn’t experience this scenario, counseling her is an essential part of managing chronic disease. 

Dr. Srivastava: I agree, we must treat the natural progression of this disease, which is not great. Dr. Sastry, is there any reason to change your management approach at this follow-up? 

Dr. Sastry: I think I would continue with bevacizumab for now because it has been effective. If her disease worsened (eg, intraretinal fluid (IRF) develops or her hemorrhage or SRF worsens) then I would consider switching medications.

Dr. Srivastava: Continuing, after her third injection of bevacizumab, this patient worsened to a VA of 20/60, and her RPE rip and PED remain unresolved. If you decide to switch medications for this patient, which one would you choose?

Dr. Arepalli: I’d probably switch to aflibercept. 

Dr. Sastry: I think I’d be inclined to switch to aflibercept 2 mg.

Dr. Borkar: I would switch to aflibercept 8 mg, partly because this patient had a complication. I believe the challenge here is managing the patient burden of every 4-week injections. I would choose a newer generation agent with more durability, such as aflibercept 8 mg or faricimab,^6,7 to “soften the blow a bit” because this patient has experienced decreased vision. In this case, I’d select aflibercept 8 mg.

Dr. Srivastava: I chose to switch this patient to faricimab. I think there is no right or wrong answer to selecting either aflibercept 8 mg vs faricimab. At 4 weeks after one injection of faricimab, her PED and fluid improved, and her VA remained at 20/60. After three monthly injections of faricimab, her hemorrhage resolved, and I was able to extend her treatment interval to 6 then 8 weeks in her right eye. 

Now, the patient notices a vision change in her left eye, and her VA decreased from 20/20 to 20/30. Her OCT scan shows SRF. Which medication do you choose at this point to treat her left eye?

Dr. Sastry: If you have the option, I’d first choose faricimab. 

Dr. Arepalli: I would start with bevacizumab.

Dr. Borkar: I would use faricimab, if possible, because it was effective on her right eye. We haven’t yet determined which drug works best for each patient in personalized medicine. However, in this case, you have the benefit of knowing faricimab worked well, so I would continue with this agent.

Dr. Srivastava: I agree and also chose to start with faricimab. After one injection, the OCT scan of her left eye shows slight improvement. After her second injection of faricimab, her OCT scan shows noticeable fluid resolution (Figure 3). Dr. Borkar, when considering newer drugs like faricimab, do you expect full resolution 4 weeks after the first injection? What are your thoughts on how quickly results should appear?

Figure 3. OCT imaging of left eye with VA of 20/30 and SRF (A), s/p faricimab x 1 (B), and s/p faricimab x 2 (C). 

Dr. Borkar: It’s ideal if the fluid resolves within 4 weeks. However, if there isn’t complete fluid resolution, I think the drug may still be working because the choroidal neovascular membrane (CNVM) may be less active. Seeing less fluid is reassuring, so encourage and reassure the patient as well.

Dr. Srivastava: Dr. Arepalli, do you have a certain number of every 4-week injections you start with no matter which agent you use? 

Dr. Arepalli: Given our increasing understanding of these new drugs, I adopt a more systematic approach, typically considering at least three injections before considering a medication switch. I believe there is merit in both methodologies.

Dr. Sastry: I usually give two or three injections before switching medications. If their retinal fluid improves with each injection, then I continue with the medication. However, if their retinal fluid worsens after two injections or there’s no improvement after three injections, I consider switching medications. 

Dr. Srivastava: With a final VA of 20/50 in her right eye and 20/30 in her left eye, this patient is responding well to an injection of faricimab every 6 weeks. Though, unfortunately, she requires every 6-week injection therapy. 

ROUND 1 | CASE 2: DETERMINING THE CAUSE OF AN RVO IN A YOUNG PATIENT

Dr. Srivastava: Our second case is a 41-year-old woman who presented with some visual distortion in her right eye. Her VA was 20/20 OU. She was healthy with no past medical history. Her medications included omeprazole and fluticasone propionate prn. She visited a general ophthalmologist who didn’t have access to an OCT but provided her a fundus photo and visual field test (Figure 4). Her left eye fundus photo and visual field test were normal.

Figure 4. Fundus photo and Humphrey’s visual field test of right eye at presentation. 

Dr. Sastry: The fundus photo of her right eye showed large blot hemorrhages, an arteriovenous crossing, and nerve fiber layer hemorrhages, all indicating a retinal vein occlusion (RVO).

Dr. Srivastava: I agree. How soon would you want to see this patient? 

Dr. Sastry: I’d want to see this patient within the next few days. 

Dr. Srivastava: This patient visits the clinic 3 weeks later (Figure 5) and, in the meantime, her ophthalmologist ordered a hypercoagulable workup. Dr. Borkar, do you typically order this type of workup for a young patient who presents with an RVO? Or what is your protocol?

Figure 5. OCT imaging of right eye 3 weeks after initial presentation. 

Dr. Borkar: For patients under 55 years of age with an RVO, I would absolutely consider a hypercoagulable workup, which includes several coagulation tests like Factor V assay, in addition to a complete blood count (CBC).⁸ I’d check the patient’s blood pressure. I do not always order an MRI of the brain and orbits in cases of RVO. 

Dr. Srivastava: I agree. Her blood pressure and CBC results were normal. Her hypercoagulable workup was negative. Dr. Arepalli, should we consider uveitis or another inflammatory disease in a 41-year-old with this presentation? 

Dr. Arepalli: I’m biased because I see a lot of misdiagnosed RVOs and retinal arterial occlusions that are unrelated to uveitis. If the retinal veins are mostly affected, then I consider inflammatory diseases, such as systemic lupus erythema (SLE) or sarcoidosis. I consider uveitis when inflammatory cells or other ocular signs are present. With an RVO, I obtain an FA to determine whether it’s ischemic or nonischemic, and whether there’s a vasculitic component. For this patient, I would also question if she had any additional risk factors, such as pregnancy?

Dr. Sastry: For a young patient with an RVO, I would also want to check for syphilis and a recent COVID-19 infection or its symptoms. 

Dr. Srivastava: Great question regarding COVID; however, her initial presentation was about 10 years ago, so she definitely had not been infected with COVID-19. Because this patient is a young woman, are there any other questions you have, including medications used, particularly contraceptives? Dr. Borkar, does the OCT show any pathology that indicates an FA is necessary?

Dr. Borkar: I see some paracentral acute middle maculopathy (PAMM) lesions, which are associated with ischemic maculopathy.⁹ So, I would obtain an FA to find out whether this RVO is ischemic or nonischemic, like Dr. Arepalli said. An FA may be helpful in determining the definitive cause of this patient’s RVO. 

Dr. Srivastava: Does observing PAMM on fundus photos and OCT scans change your management approach? 

Dr. Arepalli: No, not when the diagnosis is an RVO.

Dr. Srivastava: We obtained an FA, and her FA photo in the transit phase at 36 seconds showed excellent perfusion, including her macula, and no delay in venous filling. Her FA photo in the late phase at approximately 6 minutes showed extensive vascular leakage along the veins (Figure 6). Do you think this amount of leakage is pathognomonic for uveitis?

Figure 6. FA photos of right eye in early (A), transit (B), and late phase (C).

Dr. Sastry: Late fluorescein leakage is concerning for retinal vasculitis.¹⁰ I would initiate a vasculitis workup by ordering additional infectious and inflammatory lab workups. 

Dr. Borkar: I agree that her FA showed significant leakage. I would also order a vasculitis workup because our patient is a 41-year-old without an identified cause of her RVO. If our patient instead was a 75-year-old, then I may consider that the pathology is an idiopathic RVO. 

Dr. Arepalli: I agree as well. Again, I would look for lab results indicating diseases, such as SLE, sarcoidosis, syphilis, etc.

Dr. Srivastava: Although her VA remained at 20/20, her OCT obtained at this appointment showed more macular edema, and her fundus photo showed an increase in retinal hemorrhaging (Figure7). Her inflammatory workup was negative. Dr. Sastry, do you treat patients with minimal macular edema secondary to an RVO and a VA of 20/20 at this point?

Figure 7. OCT imaging of right eye ?? weeks later (A) and the progression analysis for the right eye (B). 

Dr. Sastry: I would not treat this patient, unless her edema was central-involving, and her visual acuity had decreased. However, I would monitor her closely and see this patient in 1 month to ensure there is no disease progression. Given the excellent perfusion of her retina, I’m less concerned about neovascularization of the retinal inner segment but, again, careful follow-up is needed.

Dr. Srivastava: So, this patient was under the care of one of my colleagues who injected her with intravitreal ranibizumab to treat her macular edema secondary to RVO.¹¹ Her OCT 2 weeks after her injection showed improvement in her macular edema and PAMM lesions. Her VA remained at 20/20. Is it helpful to know that she’s improving after one ranibizumab injection?

Dr. Borkar: I’m reassured that she’s improving, but I’m worried that I may be missing an underlying cause or systemic issue that could affect her other eye. So, I’m not entirely relieved. It’s good that the ranibizumab appears to be effective, but her edema may have resolved on its own. I would not necessarily have given her an injection at the onset of an RVO with good vision.

Dr. Srivastava: Actually, her VA did decline to 20/40, then the other ophthalmologist decided to administer an anti-VEGF injection. I question whether patients with good foveal contour, despite some edema, have definitive macular edema. How would you counsel this patient on her injections and follow-up care? 

Dr. Arepalli: She would probably not require monthly injections; however, I’d monitor her closely at first, given that her vision declined and her edema returned initially. Because she had minimal edema, I’m not inclined to obtain FAs serially. I would explain that she responded beautifully to one injection of ranibizumab so that we can hold off for now on another injection and that she should return for a follow-up in 4 to 8 weeks. 

Dr. Srivastava: At 2 and 5 months after her one ranibizumab injection, our patient continued to improve (Figure 8). Her VA was stable at 20/20, and she said, “I have a little distortion, but I’m doing pretty well.” All her workups had been negative. Her hematologist started her on daily aspirin. How do you counsel this patient? Should she be concerned about her eyes 10 years from now?

Figure 8. OCT imaging of right eye 2 months and 5 months after intravitreal ranibizumab x 1.

Dr. Sastry: Unfortunately, we do not have a definitive diagnosis, so it’s difficult to predict her future prognosis. After one injection, her retina reperfused; therefore, her RVO may have self-resolved. However, I’m inclined to continue to observe her every 6 months with an “observe and extend” approach up to 1 year.

Dr. Srivastava: Dr. Borkar, is there a certain treatment interval that indicates you can discontinue injections for patients with nonischemic RVOs and resolved edema? 

Dr. Borkar: For these patients, when their treatment interval reaches 10 to 12 weeks, I discontinue their injections and observe. I also like the “observe and extend” approach. After stopping injections, I have the patient return in about 6 weeks, then I extend the interval rapidly to 14, 16, and 18 weeks. I treat and extend differently for patients with RVOs by not loading them typically. If they respond well to the first injection, I’ll extend more quickly with RVO patients than with AMD patients.

Dr. Srivastava: Nine months after her ranibizumab injection, her fundus photo showed trace hemorrhage, and FA showed resolved vascular leakage. What do you think led to her disease improvement?

Dr. Arepalli: I don’t know if it was the ranibizumab or her natural history. However, what is striking to me is how “hot” her blood vessels appeared on her original fundus photo, which made us look for an inflammatory etiology. Because all her workups were negative for potential inflammatory diseases, I would say she has good luck and genes. 

Dr. Srivastava: For the past decade, I have seen this patient for regular follow-up. I examined her recently, and her right eye is healthy with a VA of 20/20. She remains in excellent health without any medications. She continues to ask what caused her RVO and, honestly, I don’t know. I find it challenging to determine the underlying cause of an RVO, especially nonischemic, in young patients.¹² However, they usually respond well to one anti-VEGF injection if needed. Some of these patients have rare disease courses, but her history has not suggested any systemic disease.

ROUND 1 | CASE 4: A 45-YEAR-OLD FEMALE WITH LOSS OF EZ

Dr. Srivastava: Our next patient is a 45-year-old woman with a VA of 20/20 OU. Before I say why she presented for an examination, I want to show you her OCT images for each eye (Figure 9). Can you describe what you see on the OCT and explain what features may give you concern?

Figure 9. OCT imaging of right and left eyes at presentation.

Dr. Sastry: I am noticing in the center of both scans a “flying saucer” configuration, which presents as peripheral attenuation of the external limiting membrane and ellipsoid zone (EZ) band with central preservation of these structures. My first concern with this patient is the potential of a hydroxychloroquine toxicity.

Dr. Srivastava: Other than a possible medication side effect, is there anything else to consider with loss of EZ in a young patient?

Dr. Sastry: Yes, I’d consider an inherited retinal degeneration (IRD). 

Dr. Srivastava: Yes, when a young patient presents with EZ loss consider medication toxicity (particularly hydroxychloroquine), IRD, syphilis, and trauma/retinal detachment.^13-15 For EZ loss and chronic cystoid macular edema (CME), consider birdshot chorioretinopathy, which may be accompanied by retinal vasculitis and uveitis.¹⁷

Dr. Arepalli: Interestingly, her EZ loss is “clean looking” and symmetrical.

Dr. Srivastava: Yes, she may have had syphilis previously but is presenting late. It is important to recognize EZ loss because sometimes all we obtain is an OCT scan. She had a history of SLE. She had been taking hydroxychloroquine 400 mg for the past 7 years, and her weight was 132 lbs. She was asymptomatic and had a VA of 20/20 OU. She presented to our clinic for a hydroxychloroquine screening. How do you discuss this with rheumatology? 

Dr. Arepalli: Being one of the few medications proven to reduce mortality in patients with SLE, hydroxychloroquine presents us with a challenge because of its potential for significant side effects.¹⁷ I explain to patients that its continued use can result in the loss of valuable visual field, from damage to their EZ.¹⁸ While maintaining health and staying alive are priorities for patients with SLE, preserving vision is also important. When presented with evidence on the retinal scans, this perspective often encourages rheumatologists to be more receptive to changing the treatment plan.

Dr. Borkar: These conversations with our rheumatology colleagues are quite tough, especially when the patient is on too high of a dose, like in this case. The American Academy of Ophthalmology’s guidelines for dosing of hydroxychloroquine recommend a maximum daily dose of 5 mg/kg.¹⁹ I find the conversation with a rheumatologist is easier once retinal toxicity has occurred. However, when a patient without retinal toxicity is at high risk for adverse effects, then it’s a challenging conversation to ask the rheumatologist to reduce the dosage because they see it only as a hypothetical risk.

Dr. Sastry: To best educate the patient and avoid “throwing their rheumatologist under the bus,” I say, “There’s evidence of toxicity so we want to ideally stop the hydroxychloroquine and switch to another medication, if possible. You need to coordinate with your rheumatologist to ensure that we’re getting adequate control of your lupus with an alternative medication.” 

It is important to inform them that the toxicity can persist for a period of time even after discontinuing the medication.²⁰ Therefore, even if there are no immediate visual symptoms, it is possible to experience these symptoms months or years after stopping the medication. This does not indicate inappropriate disease management; it is simply how the medication works, unfortunately. 

Dr. Srivastava: Do you have a good sense for how long these changes can continue for? 

Dr. Arepalli: No, I do not. In my fellowship, I saw a patient who developed CME and EZ loss, and they’d discontinued the medication many years prior. 

Dr. Srivastava: Unfortunately, the retinopathy screening rate in patients with SLE receiving hydroxychloroquine is low at about 23%.²¹ Dr. Borkar, do you have recommendations from your institution to ensure that patients on hydroxychloroquine have a retinal screening to avoid a missed diagnosis, such as in this case? 

Dr. Borkar: We tried adding Best Practice Advisories (BPAs) into the EHR for the rheumatology clinic; however, clinicians found the BPAs annoying. One strategy that has helped is communicating with the rheumatologist about their patients who are having retinal screenings because it reminds them to recommend a retinal screening for every patient starting on hydroxychloroquine as a baseline exam, and then for continual follow-ups.¹⁹

Dr. Srivastava: Here’s her fundus autofluorescence (FAF) and OCT imaging 8 years after discontinuing hydroxychloroquine, and she remained at 20/20 (Figure 10). What do you see here?

Figure 10. FAF and OCT images of right eye 8 years after discontinuing hydroxychloroquine.

Dr. Sastry: On the near infrared, I see a hypopigmented lesion inferotemporal to the fovea or is that artifact?

Dr. Srivastava: No, that’s a true hypopigmented lesion. So, unfortunately this patient was experiencing progressive loss of retinal tissue. Is this common? Dr. Arepalli, you’re nodding yes, while Dr. Borkar and Dr. Sastry look unsure. Do you think this warrants further investigation for other causes or diseases? 

Dr. Arepalli: Does she have any systemic conditions, such as kidney issues?

Dr. Srivastava: She doesn’t. She was only taking an immunosuppressive drug, mycophenolate. Her SLE has been relatively stable. Here she was 9 years later from the baseline of discontinuing hydroxychloroquine, and she now had a decreased VA of 20/25 OU (Figure 11). Her OCT and FAF imaging indicated early CME changes. Could these changes be attributed to other causes? Should additional diagnostic measures or imaging be considered? Are these changes typical with hydroxychloroquine usage?

Figure 11. FAF and OCT images of right and left eye 9 years after discontinuing hydroxychloroquine.

Dr. Arepalli: I’m not sure it’s common. Given the worsening pathology on FAF imaging, I would obtain a baseline workup to rule out conditions like syphilis. 

Dr. Borkar: I may get an FA for her CME. 

Dr. Srivastava: Ten years later, an FA was ordered for both eyes; this patient was now 55 years of age (Figure 12). What do you observe on her FA?

Figure 12. FA photos of early and late phases of right and left eye 10 years after discontinuing hydroxychloroquine. 

Dr. Sastry: I don’t see any leakage on the FA photos. Is this finding possibly due to an IRD, such as Goldmann-Favre syndrome?

Dr. Arepalli: Also, is hydroxychloroquine toxicity one of the factors that causes nonleaking CME on FA?

Dr. Srivastava: Yes, hydroxychloroquine toxicity and Goldmann-Favre syndrome can lead to nonleaking CME.^22,23 Our patient had a workup for IRDs, which was negative. Unfortunately, there are limited treatment options available for this young woman. Thankfully, she has maintained a VA of 20/25 OU, despite chronic CME. This case reminds us of the potential for significant retinal toxicity from the use of hydroxychloroquine.

ROUND 1 | CASE 5: ADDRESSING NONADHERENCE WITH PRP VS ANTI-VEGF THERAPY FOR PDR

Dr. Srivastava: A man in his 50s with type 2 diabetes and a HbA1c of ~12 presented with 2 weeks of vision loss. His VA was 20/400 OD and 20/20 OS. We observed no neovascularization of the iris (NVI) on slit lamp examination. His right eye showed neovascularization elsewhere (NVE), neovascularization of the optic disc (NVD), and vitreous hemorrhaging on fundus photos and capillary nonperfusion on FA (Figure 13). His left eye presented with some NVE and probable NVD. Our diagnosis is obviously proliferative diabetic retinopathy (PDR) without NVI in both eyes. How would you treat this patient?

Figure 13. Fundus and FA photos of right and left eye at presentation.

Dr. Borkar: I’m quite aggressive in my approach. This patient presented now, not years ago, for a diabetic examination because of significant symptoms. So, what are the chances he will follow-up for monthly injections? 

My other concern is the small subhyaloid hemorrhage in his right eye. When my view of the periphery is poor from significant vitreous hemorrhage inferiorly making me unable to perform good 360° panretinal photocoagulation (PRP), then I prefer to inject the patient with anti-VEGF therapy. This approach can help to avoid developing fibrovascular proliferation (FVP) and a tractional retinal detachment (TRD) if the patient with a subhyaloid hemorrhage is lost to follow-up.

Because this patient has a HbA1c of 12, I would schedule them for surgery within a few weeks. His blood glucose is currently unstable, risking high glucose levels on surgery day, which could cancel his surgery. I think that treating with anti-VEGF therapy is safe for this level of PDR; however, I avoid injections for patients with high HbA1c levels who have more advanced PDR cases. 

My plan for this patient’s surgery is to relieve all anteroposterior traction in his right eye and perform a thorough endolaser vitrectomy. While he is asleep, I would also perform a complete PRP in his other eye. I prefer not to take chances when I have a major concern of the patient being lost to follow-up. 

Dr. Arepalli: I consider results from major studies, specifically Diabetic Retinopathy Clinical Research (DRCR) Network Protocol S, which showed that patients treated with either ranibizumab or PRP had good visual outcomes at 5 years.²⁴ What worries me is that patients who receive a new diagnosis requiring extensive treatment can be overwhelmed easily, especially if the therapy is painful, eg, PRP. In clinic, I weigh the decision whether to perform PRP on both eyes in one visit or to give one anti-VEGF injection only and risk the patient not returning for follow-up. I am swayed to administer bilateral injections because they are easier to give and for the patient to receive; nonetheless, PRP may be more effective. Sometimes, I opt for PRP in both eyes and delay surgery, or I only recommend PRP to manage their PDR.

Dr. Srivastava: So, your limitation is a concern of losing rapport with the patient from the discomfort of laser treatment?

Dr. Arepalli: Yes, oftentimes patients who are treated with PRP return 2 years later with more advanced disease because they don’t want to be treated again with PRP.

Dr. Sastry: We see quite a few patients with a similar presentation to this patient. My training says to “get this patient in for PRP yesterday.” However, the art of medicine involves assessing how likely it is that the patient will return or become a “flight” risk. Recent evidence from a retrospective cohort study using the TriNetX electronic record database compared the outcomes of patients with PDR treated with PRP vs anti-VEGF injections followed by PRP.²⁵ The study found that patients who underwent PRP before receiving injections were more likely to require a vitrectomy for vitreous hemorrhage or a TRD than patients treated with anti-VEGF injections first then PRP.²⁵ Also, the DRCR Protocol AB study comparing initial treatment with anti-VEGF therapy vs vitrectomy with PRP in patients with a vitreous hemorrhage from PDR found no statistically significant difference in long-term visual outcomes between patients who received either treatment.²⁶

Therefore, the outcomes from these studies have prompted me to reconsider my treatment approach. For patients who are likely to be compliant with follow-up visits, I’d likely administer a series of anti-VEGF injections first to reduce VEGF levels and resolve vitreous hemorrhaging, if present. Subsequently, I’d plan for staged PRPs, typically at least two sessions per eye at each treatment, or vary it as needed. 

Dr. Borkar: I think Protocol AB isn’t applicable to the real world. One, this patient likely wouldn’t qualify for the study because of their subhyaloid hemorrhage. The study excluded those at risk of developing fibrovascular proliferation quickly and, instead, included simple vitreous hemorrhages.²⁶ Two, patients in the study complied with timely follow-ups and could cross over to the other study arm if needed,²⁶ which isn’t an option in the real world. 

When treating with PRP, another factor to consider is that if a small subhyaloid hemorrhage disperses or diabetic macular edema (DME) develops, the patient’s vision may worsen. This result can lead them to believe the treatment was painful and made their vision worse. So, I think gaining patients’ trust involves many considerations.

Dr. Srivastava: I appreciate each of your different responses, which indicates there is probably not a clear answer for patients with early PDR. We also tend to base our approaches off our varied patient experiences. 

Continuing, we administered an intravitreal anti-VEGF injection in the right eye, and my partner performed PRP with anesthesia at a later time. This patient was scheduled to return 3 weeks later but instead returned 9 months later. Is it common for patients not to return for follow-up as scheduled? What do you think is the issue in these scenarios?

Dr. Arepalli: Sometimes my patients’ reasoning for not returning is the pain induced by PRP. Other times, my patients with diabetes suffer from other medical conditions that require monitoring, so they don’t return because they are busy with other appointments. 

Dr. Srivastava: Patients sometimes feel better after anti-VEGF therapy, leading them to think they don’t need more laser sessions. Maybe I should have scheduled this patient for surgery initially, as Dr. Borkar suggested.

So, 9 months later, the patient had a VA of 20/200 OD and 20/20 OS with a vitreous hemorrhage in the right eye and active PDR in both eyes. What is your recommendation for this patient after a failed attempt of follow-up for laser treatment? 

Dr. Sastry: Given this patient’s history of loss to follow-up, I think that laser treatment is ultimately necessary to inactivate the disease. On the day of the procedure, you should administer an anti-VEGF injection because of the unresolved vitreous hemorrhage. If your view is adequate to administer PRP, consider performing an examination under anesthesia and completing the PRP in the operating room at that time.

Dr. Borkar: To build rapport with patients when they return late for follow-ups, I thank them for returning (despite my frustration) and emphasize that diabetes, including DME, is a chronic condition requiring ongoing treatment. I compare it to the need to remain on insulin or metformin indefinitely. I normalize pain from PRP by explaining that it can hurt regardless of the surgeon performing it. I offer two options, either receiving injections or undergoing PRP under anesthesia. For complete (not staged) PRP, I usually give a subtenon triamcinolone acetonide injection to reduce inflammation. 

Dr. Srivastava: I like your clinical thinking. So, we gave this patient three anti-VEGF injections because of his pain concerns. We agreed that I’d give him a few injections to stabilize his diabetic disease and improve his vision before intervening with laser treatment. Naturally, this patient returned a year later after his third injection (Figure 14). His right eye had a VA of 20/50 (we did perform more laser at one session); his left eye had a VA of counting fingers. Dr. Arepalli, should I recommend surgery now and finish the laser treatment? What would you do?

Figure 14. Fundus photo and B-scan ultrasound of right eye with VA of 20/50 at 21 months after presentation.

Dr. Arepalli: I’m tempted to take him to surgery. Does he tell you why he’s not coming back?

Dr. Srivastava: Great question, he had multiple reasons. He did say he felt his eyes were doing fine, so that he didn’t think he needed to come back.

Dr. Arepalli: As you mentioned, sometimes we give an anti-VEGF injection, which makes them believe they are cured. Perhaps a definitive approach, ie, PRP, could eliminate his nonadherence to follow-up.

Dr. Borkar: Again, I would have performed PRP initially in both eyes. 

Dr. Sastry: I always discuss treatment options, and their benefits and risks, with the patient. I always involve the patient in the decision-making process, although I may have a preferred treatment. My concern for PRP with this patient is that we don’t know the extent of his neovascularization, which could complicate the procedure; therefore, I’d recommend a preoperative anti-VEGF injection.

Dr. Srivastava: Dr. Borkar, are you concerned about postoperative follow-ups for this patient, given their history of loss to follow-up? 

Dr. Borkar: Yes, this patient typically returns only when he has ocular symptoms. I do concern myself with patient adherence to follow-ups; however, I’ve had patients who skip postoperative visits and eyedrop instructions, then come back years later saying they were fine and felt they didn’t need to come in. 

Dr. Srivastava: Perhaps your patients fared better because you chose a more definitive treatment initially. So, this patient agreed to anti-VEGF injections and PRP in both eyes, with a partial PRP in the right eye because of the vitreous hemorrhage. Subsequently, his VA improved to 20/40 OD and 20/50 OS, and then he developed cataracts. The patient underwent vitrectomy and cataract surgery in both eyes. We considered laser treatment, but the patient reacted with significant anxiety related to his initial laser procedure. Ultimately, he did well postsurgery reaching a VA of 20/20 OU, and we managed to control his disease effectively. Is it appropriate at this point to refer our patient to general ophthalmology for follow-up care? Should there be any concerns regarding future visits or follow-ups?

Dr. Arepalli: I’d continue to monitor him closely. Despite good PRP results, combined phacovitrectomy in patients with PDR increases the risk of anterior neovascularization.²⁷ Thus, I’d schedule him to return every 3 to 6 months initially to check for any hemorrhaging or other secondary conditions, then extend his follow-ups out. 

Dr. Srivastava: Yes, I am also concerned about DME developing. 

ROUND 2 | CASE 1: POSTINJECTION ENDOPHTHALMITIS VS RETINAL VASCULITIS

Dr. Srivastava: For the next four cases, I’m joined by Durga Borkar, MD, MMCi; Hong-Uyen Hua, MD; and Abdul-Hadi Kaakour, MD, MS. Our next case is an 80-year-old man who presented with a history of exudative AMD and postinjection endophthalmitis for 3 days. He had received his first aflibercept injection 5 days prior after 5 years of treatment with off-label bevacizumab injections given every 8 weeks. Three days ago, he began to experience severe pain, redness, and decreased vision. He went to an outside practice where he received an intravitreal injection of vancomycin, ceftazidime, and dexamethasone; no vitreous tap was performed. He was referred to us 3 days later for no improvement. How do you determine if his response is infectious or inflammatory? 

Hong-Uyen Hua, MD: First, I check for anterior chamber (AC) cells and hypopyon. While a sterile hypopyon can result from inflammatory reactions or vasculitis, it usually indicates endophthalmitis.²⁸ FA is also important here to examine the blood vessels. I compare the results in each eye, and I also assess the degree of pain. The presence of hypopyon and pain points me toward an infectious cause, and I would also do an in-office vitreous tap to evaluate for bacterial growth. 

Dr. Srivastava: Dr. Borkar, what are your thoughts on injecting without a tap? Does a tap matter if you’re going to inject vancomycin or ceftazidime?

Dr. Borkar: That’s a good question. Although it’s not my standard practice, performing a tap is quite common. In my fellowship, many private groups didn’t have lab access to analyze their sample. Besides, most of these taps are negative.²⁹ Injecting vancomycin and ceftazidime intravitreally often helps patients improve, although it’s challenging when they don’t. I’m surprised dexamethasone was injected the same day, given the suspicion for an infectious cause. For suspected infectious endophthalmitis, I typically wait 24 hours after injecting antibiotics to inject steroids.

Dr. Srivastava: Dr. Kaakour, what’s your next step in management with this patient?

Abdul-Hadi Kaakour, MD, MS: I think the patient’s language choice is very telling. He mentioned that he was not feeling better. The relevant question is whether his condition has worsened, determined partly by assessing the patient’s vision and pain levels. If a patient’s condition deteriorates after 3 days of antibiotic treatment, it raises my concern about alternative diagnoses. For example, I will question whether the patient’s condition is more inflammatory, requiring additional management, such as oral steroids. After 3 days, it is unlikely for infectious endophthalmitis to significantly improve. 

We now need to address and manage this patient’s significant inflammation, which can be destructive to intraocular tissues. So, I would consider adding topical and oral steroids to help this patient improve and manage his inflammation effectively. This approach could help control the bacterial load and reduce damage to his intraocular tissues.

Dr. Srivastava: Dr. Hua, if 3 days pass and the patient is not better or worse, but with perhaps a bit more pain, when would you consider a re-tap and re-injection? 

Dr. Hua: I have a high threshold to re-tap and inject again, instead I would consider other possible infections, such as a rare fungal endophthalmitis. I would ask the patient about pain improvement the day after the antibiotic injection. If there is no improvement, another intravitreal antibiotic injection is unlikely to help. I would then consider beginning topical steroids or increasing their dosing frequency, as well as monitor the patient closely for any worsening of signs, which indicates an infectious cause. 

Dr. Srivastava: Dr. Borkar, in cases where patients with a suspected infection aren’t worsening but not improving significantly either, what is your threshold for considering a vitrectomy? How long do you wait before intervening?

Dr. Borkar: One key test to review here is the ultrasound, particularly a B-scan, to check for increased membrane formation or signs of an abscess. A worsening of any of these findings on the B-scan ultrasound, in addition to persistent pain after 3 days (which is longer than usual), would prompt me to perform a vitrectomy.

Dr. Srivastava: Dr. Kaakour, are you comfortable prescribing oral steroids for patients, or do you prefer serial dexamethasone injections every 3 days? Or do you recommend difluprednate 0.05% ophthalmic suspension? What is your threshold for these treatments? 

Dr. Kaakour: Because this patient is of older age, I would start oral steroids cautiously and at a higher dose, assuming there are no contraindications, such as diabetes. I’m comfortable starting oral steroids the same day as intravitreal antibiotics if it’s considered safe. Even for patients with diabetes, some level of steroids is beneficial. I’ve observed cases in which a patient receives a dexamethasone injection, yet their pain and swelling worsens in a couple of days. Then, adding oral or topical steroids helps manage their inflammatory response, which gives the patient better comfort. 

Dr. Srivastava: At the time he presented to our office, the patient was without pain and his VA was hand motion. He had a hypopyon of 1 mm in his pseudophakic left eye; there was no visible view of his posterior segment. His B-scan showed no retinal detachment, and the material in the vitreous was consistent with inflammatory debris and drug precipitate. Are you confident in the diagnosis, or do you suspect an infection? What would be your next step?

Dr. Hua: It’s reassuring that his pain is improving. Reporting on changes in the size of the hypopyon and the number of AC cells would help determine whether there is improvement. Obviously, intraobserver (same person) measurements and comparisons of the hypopyon are more useful than interobserver (different persons) measurements.

I would likely increase the frequency of the topical steroid (if not already done) and continue monitoring. If his pain continues to improve, then I’d probably start oral steroids. Otherwise, if there’s any sign of worsening, I would consider an alternative management approach.

Dr. Srivastava: The good news is that after increasing the dosing of prednisolone to every 1 hour while awake, this patient’s endophthalmitis began to improve. His exam was then stable for several weeks, his VA improved from hand motion to 20/100, and a moderate amount of vitreous debris remained. Six weeks later, a vitrectomy was performed, and his VA improved to 20/40. Dr. Borkar, when you have a patient who has experienced this rare event and fears further injections, how do you explain this scenario to them? Have you faced a situation where a patient refused more injections after having endophthalmitis? 

Dr. Borkar: I’ve had patients who refused injections altogether or wanted to extend their dosing interval. For example, I had a patient on every 8-week dosing of aflibercept who wanted to extend their dosing intervals. However, once they noticed the impact on their vision from the extended interval, then I was able to convince them to resume every-8-week injections. Most patients who understand the rarity of complications are willing to continue treatment, especially with a good VA of 20/40, like this patient.

Dr. Srivastava: Yes, particularly when considering this patient’s VA was hand motion at the worst of his disease. Dr. Kaakour, how do you teach your residents and fellows to differentiate between endophthalmitis and “scary” vasculitis induced by medication?  

Dr. Kaakour: This differentiation can be very challenging because hemorrhagic, occlusive retinal vasculitis can be a late sign of endophthalmitis or an adverse event of treatments, such as vancomycin.³⁰ FA may also be useful; however, its results may be complicated when ocular inflammation leads to vascular leakage. You should also consider the timing of onset of the symptoms to help differentiate between endophthalmitis and retinal vasculitis. However, when a patient presents a few days after onset, it becomes difficult to interpret the signs accurately.

Dr. Srivastava: I recommend referring to a 2021 article by Baumal et al that summarizes proposed recommendations from an expert medical panel on managing intraocular inflammation (IOI), retinal vasculitis, and retinal vascular occlusive events that was published a few years ago in response to the rare events of IOI secondary to brolucizumab injections.³¹ For endophthalmitis, I recommend overtreating this condition. I personally foresee that we will encounter endophthalmitis and retinal vasculitis more frequently in the coming years. Thus, we need to be diligent in diagnosing these conditions quickly and treating appropriately.

ROUND 2 | CASE 2: TREATING PDR WITH ANTI-VEGF THERAPY VS PRP IN AN APPREHENSIVE PATIENT

Dr. Srivastava: A 49-year-old man with type 2 diabetes presented for a diabetic examination. His fundus photos showed scattered dot-blot hemorrhages in both eyes and an optic disc and a subhyaloid hemorrhage in the right eye. Significant findings in both eyes included leakage and mild to moderate ischemia on FA and no evidence of retinal elevation or thickening on OCT (Figure 15). His VA was 20/40 OD and 20/20 OS. We diagnosed him with PDR. Of note, this patient knew someone who went blind after a laser procedure, causing him high anxiety to the procedure. Dr. Borkar, considering the long-term data on PDR vs anti-VEGF treatment,^24,32 how do you discuss your treatment approach for PDR to a patient with hesitation to laser?

Figure 15. Fundus photo, OCT scan, and FA photo of right eye and left eye at presentation.

Dr. Borkar: The long-term data convinces me that we should use PRP for this patient, especially for their right eye. For example, in the DRCR Protocol S comparing anti-VEGF therapy to PRP in patients with PDR the rate of loss to follow-up was high at 5 years.²⁴ So, if you inject this patient in their right eye, they need to understand that injections are ongoing and that stopping injections will cause their neovascularization to return. For their left eye, injection may be an option, but I’m concerned about their compliance, given that this is their first visit to an ophthalmologist. I would recommend laser treatment and have a low threshold for an examination under anesthesia and for laser under anesthesia because this patient is anxious. 

Dr. Srivastava: Dr. Kaakour, what is your approach to treating patients with this presentation? Do you use anti-VEGF therapy for a long period of time or treat with PRP only? 

Dr. Kaakour: I’m an “old school trials” clinician. I consider that the landmark Early Treatment Diabetic Retinopathy Study and Diabetic Retinopathy Study showed PRP can reduce the risk of severe vision loss by 50% in patients with PDR.^33,34 Also, patients with PDR are more likely to have other health issues, such as kidney, heart, or brain disease, that require hospitalization, resulting in missed eye appointments for anti-VEGF treatment. PRP treatment can reduce the need for frequent follow-ups. From my clinical experience, PRP often saves more vision than it reduces. If you are concerned about contrast sensitivity loss from PRP, then guide the laser during the procedure. So, I would treat this patient with some PRP.

Dr. Srivastava: Let’s consider a scenario where a patient has had excellent PRP treatment in their right eye but experiences peripheral field loss and night vision changes. Now, their better-seeing left eye requires PRP, and they are worried about similar side effects that could affect their ability to drive. How do you address patient concerns about potential PRP complications when we have effective treatments that do not damage the retina like PRP? 

Dr. Hua: It is important to tailor treatment to each patient. If this patient was highly observant to his vision, such as pilots, then I’d strongly consider anti-VEGF monotherapy for his other eye to preserve his peripheral visual field. However, most of my patients receive a combination of treatments. My ultimate goal is to administer PRP for all eyes with PDR and inform patients about their disease. 

Building rapport with patients can take time, so I sometimes start with anti-VEGF monotherapy. I explain Protocol S showed worse outcomes for patients lost to follow-up on anti-VEGF therapy alone,²⁴ and that laser therapy is usually part of long-term management of PDR. Therefore, I begin with anti-VEGF therapy only after emphasizing the importance of consistent follow-up. Ultimately, I would recommend laser therapy for any family member with PDR.  

Dr. Borkar: Listening to patients’ fears is helpful in setting expectations. Patients may not know what type of laser treatment someone else had that resulted in vision loss. For this patient, a comprehensive PRP in the right eye with a subhyaloid hemorrhage could result in dispersed hemorrhage and decreased vision, causing this patient a similar result to the person they know. Therefore, providing anticipatory guidance about what to expect is important in managing these cases and maintaining trust over the long term. 

Dr. Srivastava: All excellent points. These cases are challenging, and patients remain ours for years, even when they are lost to follow-up for some time. An ongoing question for all is how do we ensure that patients, even those who have a history of good adherence, continue to follow-up?  

Dr. Kaakour: Although obtaining an FA isn’t necessary because PDR is often seen in the color fundus photos alone, an FA image can be a valuable tool for patients to visualize the ischemic areas of their retina. I then explain that the ischemic areas are not functioning and have already lost vision.  

Dr. Srivastava: Let me ask about the 5-year follow-up data from Protocol S that indicated patients receiving over 5 years of anti-VEGF treatment (average of 19.2 intravitreal ranibizumab injections) compared to PRP (average of 5.4 treatments) had the same vision (mean VA of 20/25 in both groups) and similar peripheral field loss (-330 dB for ranibizumab and -527 dB for PRP) at 5 years.²⁴ Additionally, the cumulative retinal detachment rate was lower in the ranibizumab group (7%) than in the PRP group (18%).²⁴ Dr. Hua, do you believe the visual field finding from this trial data, considering the sample size was small? 

Dr. Hua: Yes, the study cohort was small. Patients develop PDR due to peripheral ischemia, and anti-VEGF injections help to suppress the VEGF drive but do not improve ischemia; therefore, I wouldn’t expect the visual field loss to be significantly less in the patients treated with anti-VEGF therapy vs PRP.

Dr. Srivastava: Have you had any patients receive 5 years of anti-VEGF injections only for PDR? 

Dr. Borkar: No, I usually deliver combination therapy for PDR.  

Dr. Kaakour: I’ve had patients who received anti-VEGF treatment for years then discontinued treatment when their HbA1c improved. However, after some years, signs of diabetic retinopathy (DR) returned, requiring treatment again. 

Dr. Srivastava: Continuing with our patient, almost 2 years later he had a VA of 20/25 OD after two sessions of PRP and five injections of anti-VEGF therapy for DME (Figure 16). His left eye had a VA of 20/25 after treatment with two anti-VEGF injections only. The patient had missed two follow-up appointments, leading to concerns about the future trajectory of his left eye.

Figure 16. Fundus photo and OCT scan of right eye (s/p PRP x 2 and anti-VEGF injection x 5) and left eye (s/p anti-VEGF injection x 2) 20 months after presentation.

I also want to discuss a related case of a 51-year-old man presenting with blurred vision of 2 months in his left eye. His IOP was 21 mm Hg OD and 52 mm Hg OS. In his left eye, we observed NVI of 2+ to 3+ on slit lamp examination and ischemia, leakage, and NVD on his FA photos. In his right eye, we observed ischemia and leakage on his FA photos. How do you treat patients with a high IOP from neovascular glaucoma? 

Dr. Kaakour: Treatment for this patient is “all hands-on deck.” His left eye is at the end stage of DR, in addition to a high IOP; therefore, his vision may deteriorate quickly. Besides the need to obviously focus on treating his left eye, we need to pay attention to his right eye because of the areas of ischemia and leakage on the nerve. Although his right eye currently lacks NVI, it could develop quickly. So, I’d recommend anti-VEGF therapy and laser treatment for his left eye and anti-VEGF therapy for his right eye, potentially using laser treatment to preserve vision for as long as possible.

Dr. Srivastava: Dr. Borkar, do you refer patients with a high IOP to your glaucoma specialist quickly? Or do you treat the high IOP in his left eye before giving anti-VEGF therapy or PRP? How would you comanage this patient?

Dr. Borkar: If this patient with an IOP above 50 mm Hg presented at my clinic, I would administer an anti-VEGF injection and perform an AC tap. The patient’s cornea is likely “too steamy” or edematous to perform PRP same day. I would start this patient on antihypertensive drops and refer them to my glaucoma colleague within the week. Antihypertensive drops alone may not be as effective for these patients because their angle is often closed, which requires a tube for fluid drainage to decrease their IOP. Thus, drops provide a temporary solution until surgery can be scheduled. I think that performing PRP before tube-shunt surgery when the patient is under anesthesia is a practical approach, if logistically possible.

Dr. Srivastava: Great point, it may also be practical to treat both eyes at the same time. Dr. Hua, when you have a patient like this who is concerned about their left eye but who also needs treatment in their other eye, how and when do you address treatment in their right eye? Where does treating the right eye fit in your priorities? 

Dr. Hua: Building rapport with patients is obviously important. I’ll explain to the patient, “Although you’ve lost a lot of vision in your left eye, we’re going to try all available treatments to recover your vision. However, it’s also important to address and protect your right eye. So, I suggest injecting both eyes with anti-VEGF therapy. Our priority is to treat your better-seeing right eye also with laser therapy to preserve its vision for life.” This explanation helps to gain buy-in from the patient.  

Dr. Srivastava: Great, so this patient received laser therapy in both eyes. His right eye did not receive any anti-VEGF injections and had a VA of 20/25 after one session of PRP. Some active disease remained in his right eye; however, it responded quite well to treatment. His left eye had a VA of 20/40 after two PRP treatments, two anti-VEGF injections, and a glaucoma tube-shunt surgery.

Four years later, we treated his right eye with PRP, and his VA remained at 20/25. Unfortunately, he lost some nerve fiber layer in his left eye and remained at a VA of 20/40. Therefore, even with prompt therapy, these patients can suffer permanent ocular damage, although the visual outcomes may be fairly good considering the concerning presentation and chronic disease of their eyes.

ROUND 2 | CASE 4: ACUTE SEVERE VISION LOSS AND RETINAL WHITENING IN A 24-YEAR-OLD WOMAN

Dr. Srivastava: A 24-year-old woman presented with a 2-week history of severe vision loss in both eyes. She had been previously diagnosed with COVID-19 and, on the following day, she experienced vision loss in both eyes. She was initially examined in an emergency room 2 weeks prior, where a CTA scan of her neck returned negative results. Although the administration of tissue plasminogen activator was considered, it was not given because she was outside the window for treatment. Her blood pressure was recorded at 104/64 mm Hg. This patient had previously been vaccinated for COVID and was evaluated by an ophthalmologist who determined that she had COVID-associated vision loss. She was on oral contraceptives and taking medication for an anxiety disorder. Her VA was 20/400 OD and counting fingers OS. Her fundus photos and FAF imaging were unremarkable in both eyes. There was no evidence of vitreal cells in either eye.

In both eyes, her FA photos showed no peripheral vascular leakage or intraocular inflammation. The FA photo of her left eye showed slight leakage around a few vessels inferiorly to the optic nerve, which did not appear affected. Her OCT imaging of both eyes showed hyperreflectivity in the inner to middle macula, with the heat map reflecting this anomaly (Figure 17). So, her presentation included intraretinal and deep retinal whitening, resembling a PAMM lesion. Could this be a COVID-associated inflammatory disorder?

Figure 17. OCT scan of right eye (A) and left eye (B) at presentation.

Dr. Hua: It’s easy to associate this presentation with COVID, which is known for causing a hypercoagulable state³⁵; however, a thorough systemic workup is necessary. I would consider other possible infectious causes, specifically sarcoidosis, syphilis, and tuberculosis. After ruling out some of these potential infectious causes of her retinal findings, I may consider treating with corticosteroids. Given her PAMM lesions and hypercoagulable state, I would also order a vascular workup. So, while a COVID-associated disorder isn’t impossible, I’d conduct a comprehensive workup. 

Dr. Kaakour: I agree that a hypercoagulable workup is necessary in this case. I would also consider Susac syndrome because of her history of anxiety, hearing loss, and possible arterial occlusive disease,³⁶ as seen on her FA imaging suggests, particularly in her left eye. Note that a CTA scan may not rule out arterial occlusive disease in certain cases. Her history of oral contraceptive usage also raises my concerns of hypercoagulability.³⁷ So, this patient presented with other confounding factors, besides a COVID-19 infection.

Dr. Srivastava: Great work, we have created a complete list of potential causes of her vision loss. 

Dr. Borkar: Yes, the colluding has begun. I have no further comments to add as my colleagues have comprehensively addressed this case, in my opinion. 

Dr. Srivastava: Excellent, you were all on the right track. Let’s review my approach. Retinal whitening on OCT can be due to many causes, including infectious, inflammatory, neoplastic, therapeutic, and idiopathic.^38,39 When a patient has inner retinal whitening, I consider Behcet’s disease, syphilis, sarcoidosis, anti-neutrophil cytoplasmic antibody (ANCA)-associated vasculitis, SLE, antiphospholipid syndrome (APLS), and Bartonella henselae (commonly referred to as cat scratch disease). When presented with bilateral inner retinal whitening, my key recommendation is to always consider B. henselae,⁴⁰ which can be transmitted by many different animals and insects.⁴¹ When the inner retinal whitening is multifocal in both eyes acutely and simultaneously, I strongly consider syphilis, sarcoidosis, and B. henselae, and it’s less likely but possibly Behcet’s disease, SLE, or APLS. 

This patient had a positive serological test for B. henselae IgM (1:128) and IgG (1:64) antibodies. Despite initial skepticism of B. henselae, subsequent tests showed increasing IgG levels, confirming our diagnosis. Clinicians may say this infection doesn’t exist, but it does. My recommendation is to check the serological test more than once at about 6 to 8 weeks later, even when the results are initially negative, because the IgG levels may increase. Another recommendation I have is to be sure to consider all masqueraders of retinal disease, especially with inner retinal disease.  

After starting steroids, this patient’s vision improved to a VA of 20/30 OD and 20/200 OS 2 weeks later, and to a VA of 20/25 OD and 20/70 OS at 6 weeks. However, she had developed large, paracentral scotomas; her OCT angiography scan showed the correlating, extensive loss in both eyes (Figure 18). At 2 years, her VA reached 20/20 OU; nonetheless, this patient was aware of the scotomas in her field of vision. Although her condition improved, we were unable to resolve all her symptoms. 

Figure 18. OCT angiography scan of right and left eye at 6 weeks. 

ROUND 2 | CASE 5: A POSTOPERATIVE CATARACT SURGERY CONUNDRUM AT 1 WEEK 

Dr. Srivastava: Our last case is a 60-year-old man who had cataract surgery in his right eye, which was complicated by a posterior capsule rupture requiring an anterior vitrectomy. Topical miotic eyedrops were administered during the procedure. On postoperative day 1, he had a VA of 20/200 and an IOP of 19 mm Hg, and his IOL was centered. Sutures were present on the corneal wound. The patient had experienced some vomiting on postoperative day 0 but recovered by postoperative day 1. On postoperative day 7, the patient’s VA declined to 20/70 (pinhole 20/25), and his IOP increased to 25 mm Hg. His IOL appeared tilted with one haptic in the AC and the other haptic pointing posteriorly. The cataract surgeon observed choroidal effusions or choroidals nasally, inferiorly, and temporally, leading him to refer this patient to our clinic for a consultation. Dr. Kaakour, what are your thoughts? What do you think caused this lens to tilt after postoperative day 1?  

Dr. Kaakour: We can’t be certain whether the IOL was placed correctly in the sulcus at the time of surgery or whether it may have tilted or shifted because of the patient’s vomiting or increased IOP from a potential choroidal formation. Currently, the patient exhibits symptoms of uveitis-glaucoma-hyphema syndrome without hyphema. Although hypopyon was not documented at this point, inflammation in the eye is likely. 

Dr. Srivastava: He had trace to 1+ cells, or mild inflammation.

Dr. Kaakour: Because this patient had a better visual acuity with pinhole, I think that the choroidal effusions are likely shallow. My primary aim now is to determine the cause of his vomiting, which may be attributable to the use of topical miotic drops or anesthetic medications. I would medically manage this patient and help prevent further increases in his IOP using atropine and steroids. I’d also closely monitor his choroidal effusions for improvement. 

Dr. Borkar: Yes, it’s difficult to know the cause of his tilted IOL. I agree with Dr. Kaakour’s approach to medically manage the patient at this point. 

Dr. Hua: Perhaps, this patient’s IOL was tilted at postoperative day 1 but was not visible because of his miotic pupil. I will add that I’d schedule this patient same day for a baseline examination and imaging. In addition to the conservative medical management mentioned, I may also recommend oral steroids.  

Dr. Srivastava: Continuing, I examined this patient the next day and observed a 360° choroidal hemorrhage with a few crystalline lens fragments in the posterior vitreous. His IOP was 24 mm Hg, and his VA was 20/40. His IOL remained dislocated with one haptic in the AC and the other in the vitreous. When would you operate on this patient who was frustrated that his vision was worse?

Dr. Kaakour: I’d be comfortable monitoring this choroidal hemorrhage because it is not appositional and expected to liquefy in a few weeks. There’s also no urgency to operate immediately because his AC isn’t shallow or collapsed, and there’s no corneal-lens touch. Again, I’d manage him medically, being cautious not to lower his IOP drastically because his initial low IOP may have caused hemorrhagic choroidals, which is difficult to see without a B-scan ultrasound. I’d monitor him until the choroidal hemorrhages resolved and then reconsider intervention in a few weeks. I’ll also check whether he’s on anticoagulant medication, and if so, assess whether it can be safely discontinued to prevent possible expansion of the choroidals. 

Dr. Srivastava: Dr. Borkar, what is your potential surgical plan for this patient? 

Dr. Borkar: In this case, we were fortunate that his IOP was controlled. Again, the choroidals weren’t appositional, so we could wait 10 to 14 days for surgery. Once choroidals are liquefied, I can achieve a good drain by focusing on a couple of quadrants, generally opposite each other. I also perform a vitrectomy in these cases. Since this patient recently had an anterior vitrectomy, it’s prudent to perform a more thorough vitrectomy now. During surgery, I’d assess his sulcus and determine if it’s feasible to have a sulcus “lensless;” I have a lower threshold for the sulcus’ stability when considering removing an IOL. 

Dr. Hua: I’d take a “less is more” approach, given his serious complication of hemorrhagic choroidals. Like Dr. Borkar, I’d perform a vitrectomy, drain, and remove the IOL. I’d place another IOL at a later surgery to avoid the risk of a longer surgery for this patient.

Dr. Kaakour: I think the management plan depends on the patient’s goal. If resolving the issue by removing the IOL is the patient’s priority, then I may consider inserting an ACIOL based on his age. This approach avoids complications behind the iris. I would safely remove the IOL by creating a scleral tunnel, then insert an ACIOL and suture properly, ensuring no leaking. 

Dr. Srivastava: To summarize the surgery (Figure 19), I used an AC maintainer, and I marked my midpoint. I made partial thickness sclerotomies at approximately 3.5 mm posterior to the corneal limbus for choroidal drainage. How do you approach sclerotomies, and how do you know when to stop pressing on the posterior lip while draining the blood?

Figure 19. To view the surgery, click to link to the enduring program. The surgery begins at the 1:10:00 mark of the video.

Dr. Hua: Sometimes I use forceps or another instrument to apply gentle pressure on one lip of the partial thickness sclerotomy to allow the blood to continue flowing out. Occasionally, you can consider increasing the IOP to 60 mm Hg to facilitate further release of the suprachoroidal hemorrhage. 

Dr. Borkar: In most cases, I also perform a pars plana vitrectomy, which allows me to examine the choroidals. Once the blood is not easily coming out, I begin the procedure. When the effusion is in the vitreous cavity, I increase the IOP to 60 mm Hg for further blood release. During the vitrectomy, additional drainage of blood usually occurs, which is typically sufficient. 

Dr. Srivastava: For this patient, the three partial thickness sclerotomies drained an adequate amount of blood, making an additional procedure unnecessary. My fellow who aided in the surgery brought the IOL into the AC. We observed that the patient’s IOL is backwards and flipped upside down (ie, “S” configuration) and that his sulcus is supportive. How would you proceed? 

Dr. Kaakour: I personally don’t trust the sulcus support. I’d remove his IOL and insert an ACIOL.

Dr. Borkar: If the sulcus is adequately support as noted, I would consider keeping the current IOL. Otherwise, I would remove the flipped IOL and insert a new one. I would not insert an ACIOL because of my concern for its long-term impact on his cornea, given the patient’s relatively young age. I’d also probably not “flip” or reorient this patient’s IOL because it has been in place for so long, ie, removing and reinserting the flipped IOL would require a larger wound. I think the better approach is to create a small 2.5-mm corneal incision, remove the flipped IOL, and insert a new IOL. 

Dr. Hua: Aspheric IOLs can be placed in the flipped or “S” configuration, and the patient’s vision should not be affected theoretically. However, I’d first attempt to flip his IOL and place it in the sulcus. I believe the fewer surgical holes in the eye, the better.

Dr. Srivastava: So, we decided to flip this patient’s IOL and replace it into the sulcus. During the surgery, our fellow demonstrated a technique where she expertly grasped and pulled the IOL posteriorly in one fluid motion, placing it into each side of the sulcus using the Kuglen Hook. We essentially had to manipulate the flipped IOL from one hand to the other hand using forceps through the cannula. We used iris hooks to ensure clear visibility throughout the procedure, which highlights the importance of ensuring a full dilation. This patient reached a VA of 20/20 at 6 weeks postoperatively. Although many circumstances of this case are rare, the key lesson is to have patience and use strategic clinical decisions to achieve positive patient outcomes in challenging cases.

1. Tanaka A, Hata M, Tsuchikawa M, et al. Short-term outcomes of 3 monthly intravitreal faricimab on different subtypes of neovascular age-related macular degeneration. Clin Ophthalmol. 2024;18:507-516. 

2. Veritti D, Sarao V, Gonfiantini M, Rubinato L, Lanzetta P. Faricimab in neovascular AMD complicated by pigment epithelium detachment: an ai-assisted evaluation of early morphological changes. Ophthalmol Ther. 2024;13(10):2813-2824. 

3. Lai TYY, Khanani AM, Souverain A, Yang M, Patel S, Margaron P. Greater reduction in pigment epithelial detachment size with faricimab vs aflibercept during head-to-head dosing in patients with neovascular age-related macular degeneration. Presented at the 16th Congress of the Asia-Pacific Vitreo-Retina Society Hong Kong Convention and Exhibition Centre, Hong Kong. December 8-10, 2023. 

4. Chew EY, Clemons TE, Agrón E, et al. Long-term outcomes of adding lutein/zeaxanthin and ω-3 fatty acids to the AREDS supplements on age-related macular degeneration progression: AREDS2 Report 28. JAMA Ophthalmol. 2022;140(7):692-698. 

5. Age-Related Eye Disease Study 2 Research Group. Lutein + zeaxanthin and omega-3 fatty acids for age-related macular degeneration: the Age-Related Eye Disease Study 2 (AREDS2) randomized clinical trial. JAMA. 2013;309(19):2005-2015. 

6. Regeneron. Eylea HD Package Insert. https://www.regeneron.com/downloads/eyleahd_fpi.pdf. Revised December 2024. Accessed February 9, 2025. 

7. Genentech. Vabysmo Package Insert. https://www.gene.com/download/pdf/vabysmo_prescribing.pdf. Revised July 2024. Accessed February 9, 2025.

8. Tauqeer Z, Bracha P, McGeehan B, VanderBeek BL. Hypercoagulability testing and hypercoagulable disorders in young central retinal vein occlusion patients. Ophthalmol Retina. 2022;6(1):37-42. 

9. Chu S, Nesper PL, Soetikno BT, Bakri SJ, Fawzi AA. Projection-resolved OCT angiography of microvascular changes in paracentral acute middle maculopathy and acute macular neuroretinopathy. Invest Ophthalmol Vis Sci. 2018;59(7):2913-2922. 

10. Agarwal A, Afridi R, Agrawal R, Do DV, Gupta V, Nguyen QD. Multimodal imaging in retinal vasculitis. Ocul Immunol Inflamm. 2017;25(3):424-433. 

11. Regeneron. Lucentis Package Insert. https://www.gene.com/download/pdf/lucentis_prescribing.pdf. Revised February 2024. Accessed February 9, 2025.

12. Chen TY, Uppuluri A, Zarbin MA, Bhagat N. Risk factors for central retinal vein occlusion in young adults. Eur J Ophthalmol. 2021;31(5):2546-2555. 

13. Arepalli S, Traboulsi EI, Ehlers JP. Ellipsoid zone mapping and outer retinal assessment in stargardt disease. Retina. 2018;38(7):1427-1431. 

14. Kiss S, Damico FM, Young LH. Ocular manifestations and treatment of syphilis. Semin Ophthalmol. 2005;20(3):161-167. 

15. Figueiredo N, Sarraf D, Gunnemann F, et al. Longitudinal assessment of ellipsoid zone recovery using en face optical coherence tomography after retinal detachment repair. Am J Ophthalmol. 2022;236:212-220. 

16. Bousquet E, Duraffour P, Debillon L, Somisetty S, Monnet D, Brézin AP. Birdshot chorioretinopathy: A review. J Clin Med. 2022;11(16):4772. 

17. National Library of Medicine. Hydroxychloroquine. NLM: Medlineplus.gov; 2023. Accessed February 9, 2025. https://medlineplus.gov/druginfo/meds/a601240.html#:~:text=Hydroxychloroquine%20is%20in%20a%20class,activity%20of%20the%20immune%20system.

18. Yusuf IH, Charbel Issa P, Ahn SJ. Hydroxychloroquine-induced retinal toxicity. Front Pharmacol. 2023;14:1196783. 

19. American Academy of Ophthalmology. Recommendations on screening for chloroquine and hydroxychloroquine retinopathy-2016. March 2016. Accessed February 10, 2025. https://www.aao.org/education/clinical-statement/revised-recommendations-on-screening-chloroquine-h

20. Mititelu M, Wong BJ, Brenner M, Bryar PJ, Jampol LM, Fawzi AA. Progression of hydroxychloroquine toxic effects after drug therapy cessation: new evidence from multimodal imaging. JAMA Ophthalmol. 2013;131(9):1187-1197.

21. Lee JE, Nam DR, Sung YK, Kim YJ, Jung SY. Nationwide patterns of hydroxychloroquine dosing and monitoring of retinal toxicity in patients with systemic lupus erythematosus. Sci Rep. 2023;13(1):7270.

22. Parikh VS, Modi YS, Au A, et al. Nonleaking cystoid macular edema as a presentation of hydroxychloroquine retinal toxicity. Ophthalmology. 2016;123(3):664-666. 

23. Lingao MD, Ganesh A, Karthikeyan AS, Al Zuhaibi S, Al-Hosni A, Al Khayat A, et al. Macular cystoid spaces in patients with retinal dystrophy. Ophthalmic Genet. 2016;37:377–83. 

24. Gross JG, Glassman AR, Liu D, et al. Five-year outcomes of panretinal photocoagulation vs intravitreous ranibizumab for proliferative diabetic retinopathy: A randomized clinical trial. JAMA Ophthalmol.2018;136(10):1138–1148. 

25. Alsoudi AF, Wai KM, Koo E, Parikh R, Mruthyunjaya P, Rahimy E. Initial therapy of panretinal photocoagulation vs anti-VEGF injection for proliferative diabetic retinopathy. JAMA Ophthalmol. 2024;142(10):972-975. 

26. Antoszyk AN, Glassman AR, Beaulieu WT, et al. Effect of intravitreous aflibercept vs vitrectomy with panretinal photocoagulation on visual acuity in patients with vitreous hemorrhage from proliferative diabetic retinopathy: A randomized clinical trial. JAMA. 2020;324(23):2383–2395.

27. Hwang S, Kim JH, Choi YH, et al. Incidence and risk factor analysis of neovascular glaucoma following vitrectomy in patients with proliferative diabetic retinopathy. Sci Rep. 2024;14(1):22490. 

28. Gil-Martínez M, Rodríguez-Cid MJ, Fenández-Rodríguez MI, et al. Clinical features, management and outcomes of patients with sterile endophthalmitis associated with intravitreal injection of antivascular endothelial growth factor. Arch Soc Esp Oftalmol (Engl Ed). 2020;95(5):211-216.

29. Xu K, Chin EK, Bennett SR, et al. Endophthalmitis after intravitreal injection of vascular endothelial growth factor inhibitors: management and visual outcomes. Ophthalmology. 2018;125(8):1279-1286. 

30. Cunningham ET Jr, Moorthy RS, Witkin AJ, Zierhut M. Occlusive retinal vasculitis following intravitreal drug delivery. Ocul Immunol Inflamm. 2022;30(6):1501-1504.

31. Baumal CR, Bodaghi B, Singer M, et al. Expert opinion on management of intraocular inflammation, retinal vasculitis, and vascular occlusion after brolucizumab treatment. Ophthalmol Retina. 2021;5(6):519-527.

32. Macaron MM, Al Sabbakh N, Shami MZ, et al. Anti-VEGF injections vs. panretinal photocoagulation laser therapy for proliferative diabetic retinopathy: A systematic review and meta-analysis. Ophthalmol Retina. 2025;9(2):105-121.

33. Ferris F. Early photocoagulation in patients with either type I or type II diabetes. Trans Am Ophthalmol Soc. 1996;94:505-537. 

34. Diabetic Retinopathy Study Research Group. Indications for photocoagulation treatment of diabetic retinopathy: Diabetic Retinopathy Study report number 14. Int Ophthalmol Clin. 1987;27(4):239-253. 

35. Ranucci M, Ballotta A, Di Dedda U, et al. The procoagulant pattern of patients with COVID-19 acute respiratory distress syndrome. J Thromb Haemost. 2020;18(7):1747-1751. doi:10.1111/jth.14854

36. Bose S, Papathanasiou A, Karkhanis S, et al. Susac syndrome: neurological update (clinical features, long-term observational follow-up and management of sixteen patients). J Neurol. 2023;270(12):6193-6206.

37. Lo Faro V, Johansson T, Johansson Å. The risk of venous thromboembolism in oral contraceptive users: the role of genetic factors-a prospective cohort study of 240,000 women in the UK Biobank. Am J Obstet Gynecol. 2024;230(3):360.e1-360.e13.

38. Agarwal A, Rübsam A, Zur Bonsen L, Pichi F, Neri P, Pleyer U. A comprehensive update on retinal vasculitis: Etiologies, manifestations and treatments. J Clin Med. 2022;11(9):2525. 

39. Arepalli S, Fondriest J, Raiji V. Uveitis corner: Interpreting optical coherence tomography findings of retinal whitening. Retinal Physician. 2023;20(April):40-45. Accessed February 11, 2025. https://digital.retinalphysician.com/publication/frame.php?i=787603&p=&pn=&ver=html5&view=articleBrowser&article_id=4548162

40. Tey MS, Govindasamy G, Vendargon FM. The clinical spectrum of ocular bartonellosis: a retrospective study at a tertiary centre in Malaysia. J Ophthalmic Inflamm Infect. 2020;10(1):31. 

41. U.S. Centers for Disease Control and Prevention. Bartonella Infection. May 15, 2024. Accessed February 11, 2025. https://www.cdc.gov/bartonella/about/index.html#:~:text=Key%20points,bartonellosis%20in%20the%20United%20States.