AMD and Anti-VEGF: Trials, Dosing Regimens and High-Yield Points

The safest answer to an AMD anti-VEGF trial question has five parts: phenotype, study population, intervention/comparator, dosing language, and outcome/limitation. Do not memorise MARINA, ANCHOR, CATT, IVAN and VIEW as a bare acronym list. Their value in an examination is that each teaches a different question: registration-era efficacy against a control, comparison between agents, comparison between dosing strategies, or interval-based non-inferiority design. This is exam education for doctors, not a recommendation for a particular patient’s injection schedule.

First identify the AMD phenotype and the imaging question

Age-related macular degeneration is commonly discussed as non-neovascular (often called dry) and neovascular AMD. In a retina viva, “wet” alone is not a complete answer. Describe the lesion and the evidence: drusen/pigmentary background, subretinal or intraretinal fluid on OCT, pigment epithelial detachment or sub-RPE material where relevant, haemorrhage, and correlation with fluorescein angiography, ICGA or OCTA when supplied. The AAO AMD Preferred Practice Pattern is the appropriate current professional reference for diagnostic terminology and management context; a trial paper is not a substitute for a guideline.¹

The classic trials below largely used the terminology and imaging standards of their own era. Therefore do not force an older fluorescein-angiographic lesion category into a modern OCT answer, and do not assume that a dry OCT at one visit fully defines disease activity in every clinical setting. In an exam, state what the image shows, then state the uncertainty.

The trial-card method

For every acronym make one card with five fields:

This method is higher-yield than learning a percentage without the denominator. If a stem changes one field, you can still reason to the study family.

MARINA and ANCHOR: what the ranibizumab era established

MARINA studied ranibizumab in minimally classic or occult choroidal neovascularisation due to AMD, comparing scheduled intravitreal ranibizumab with sham injection. At two years, the paper reported better visual-acuity outcomes for the ranibizumab groups than sham.² The exam anchor is therefore: occult/minimally classic lesions; ranibizumab; sham-controlled efficacy framework. Avoid merging it with ANCHOR.

ANCHOR studied predominantly classic CNV and compared ranibizumab with verteporfin photodynamic therapy. It likewise demonstrated superior visual outcomes for ranibizumab under the trial protocol.³ The anchor is: predominantly classic lesions; ranibizumab versus verteporfin PDT. When an examiner asks why the two trial names are paired, the answer is not “both were anti-VEGF trials”; it is that they addressed different lesion classifications from that era and established efficacy against different controls.

Neither study creates an individual prognosis. Eligibility, visit schedule, imaging, adverse-event monitoring and subsequent care differ from one patient to another. State the protocol before extrapolating.

CATT and IVAN: do not confuse agent comparison with regimen comparison

The Comparison of AMD Treatments Trials (CATT) randomised participants with untreated neovascular AMD to ranibizumab or bevacizumab and to monthly or as-needed dosing assignments, with re-randomisation of the monthly groups at one year.⁴ Its first-year report found that, under its protocol, bevacizumab and ranibizumab had similar effects on visual acuity when administered on the same schedule; monthly dosing produced slightly better visual outcomes than as-needed dosing, while as-needed treatment used fewer injections.⁴ This is why CATT appears in both “agent” and “frequency” questions.

The important caveat is equally examinable. “PRN” in CATT is not an undefined instruction to inject whenever someone feels it is needed. It was protocolised: monitoring and retreatment criteria were defined by the study. Do not quote a CATT result as proof that every real-world PRN system will perform identically.

IVAN was a UK multicentre factorial trial that also compared bevacizumab with ranibizumab and continuous with discontinuous treatment strategies in neovascular AMD.⁵ It is often tested alongside CATT because it independently addresses a similar broad comparison. In a written answer, give the design idea rather than blending outcomes from the two trials. If exact numerical results are requested, return to the original publication or the paper specified by the examiner.

VIEW 1 and VIEW 2: the aflibercept interval question

The VIEW 1 and VIEW 2 studies compared intravitreal aflibercept regimens with monthly ranibizumab in neovascular AMD. A key high-yield memory point is the aflibercept 2 mg every-eight-weeks schedule after three initial monthly doses; the pivotal paper reported non-inferior visual-acuity outcomes at one year for the tested aflibercept regimens versus monthly ranibizumab under the study design.⁶ “Every eight weeks” is therefore a trial-protocol fact, not a universal instruction for every person with AMD.

Examiner wording matters. If it asks “which trial supported an eight-week aflibercept regimen after loading?”, VIEW is the expected family. If it asks “what is a loading phase?”, answer generically: a planned series of initial treatments in a specified protocol. Do not claim that every agent, label, country or contemporary regimen uses the same loading pattern.

Fixed, PRN and treat-and-extend: three phrases, three logics

Treat-and-extend is a practical management paradigm with variable protocols; it is not synonymous with the PRN arms of CATT or IVAN. It is also not merely “give fewer injections”. The concept is to maintain treatment while extending or shortening intervals according to a defined activity assessment. The exact endpoint, maximum interval and handling of fluid differ across studies and services. A strong candidate labels that variability rather than declaring a universal rule.

How to answer the OCT part of a trial stem

Use a consistent order:

1. Say which eye and whether scan quality allows interpretation.
2. Identify intraretinal fluid, subretinal fluid, sub-RPE material/PED, hyperreflective foci, fibrosis or haemorrhage if shown.
3. Relate the feature to the suspected macular neovascular process only to the extent the image supports.
4. State which additional imaging or longitudinal comparison would clarify uncertainty.
5. Separate the image description from the study evidence and from management.

For example: “The OCT shows subretinal fluid with a pigment epithelial detachment in an eye with drusen. This supports exudative activity in the appropriate clinical context; angiographic/OCTA correlation and serial scans may clarify lesion characteristics.” It is more defensible than stating “this patient needs drug X at interval Y”.

Fifteen AMD and anti-VEGF MCQs with explanations

1. MARINA mainly enrolled which lesion types? Minimally classic or occult CNV due to AMD. Why: contrast it with ANCHOR’s predominantly classic population.²
2. ANCHOR compared ranibizumab with what? Verteporfin PDT. Why: the comparator is the key discriminator.³
3. CATT compared which agents? Bevacizumab and ranibizumab. Why: it was also a dosing-frequency trial.⁴
4. CATT compared which schedules? Monthly and as-needed assignments. Why: say “under protocol-defined monitoring/retreatment”, not generic PRN.⁴
5. Why is CATT not simply a “bevacizumab trial”? It tested both agent and schedule questions. Why: factorial design changes interpretation.
6. IVAN is commonly paired with CATT because? Both compared bevacizumab/ranibizumab and treatment strategies in neovascular AMD. Why: do not merge their data.⁵
7. VIEW is most associated with? Aflibercept in neovascular AMD. Why: remember the 2 mg eight-week regimen after initial monthly doses.⁶
8. Fixed dosing means? A pre-specified interval. Why: it does not name one particular interval.
9. PRN means? Retreatment after monitoring according to defined criteria. Why: it is not “review only if symptomatic”.
10. Treat-and-extend differs from PRN because? It generally maintains proactive treatment while adjusting interval. Why: protocol details vary.
11. Drusen alone establish neovascular AMD? No. Why: look for lesion/exudative evidence and context.
12. A dry OCT at one visit proves permanent inactivity? No. Why: interpretation is longitudinal and clinical.
13. Can an old study’s lesion labels be copied uncritically into OCT language? No. Why: describe the modality and era-specific terms accurately.
14. Does trial efficacy equal an individual prognosis? No. Why: trial population and protocol bound the inference.
15. Best viva closing sentence? “I would describe the phenotype and imaging, then state the trial’s population, comparator and schedule before discussing applicability.” Why: it separates evidence from prescription.

Five harder prompts for written papers

16. Why is “loading dose” an incomplete answer? It omits agent, number, interval and protocol context.
17. Which question does a sham-controlled trial answer best? Efficacy relative to a no-active-injection control in its defined population.
18. Which issue limits simple comparison of trial injection burden? Different monitoring, retreatment and crossover rules.
19. What is non-inferiority? A design testing whether a treatment is not worse than a comparator by more than a pre-specified margin; name the margin only from the source.
20. What should you do when a stem asks exact outcome percentages? Verify the named paper and time point; do not rely on an acronym mnemonic.

Revision plan: 45 minutes, one page of notes

Make five trial cards using the table above. On the reverse, write one discriminating phrase: MARINA “occult/minimally classic versus sham”; ANCHOR “predominantly classic versus PDT”; CATT “agent plus monthly/PRN”; IVAN “UK factorial comparison”; VIEW “aflibercept interval programme”. Spend fifteen minutes on OCT descriptions from a standard retina text. Then answer the twenty questions without notes and record whether each error was population, comparator, regimen language, outcome interpretation or imaging. Revisit the error category in a mixed retina session. Do not attempt to learn trial results as a treatment algorithm.

Worked written-answer structures

“Compare MARINA and ANCHOR”

Open with what the trials shared: both were pivotal ranibizumab programmes for neovascular AMD. Then state the difference that earns marks: MARINA addressed minimally classic or occult lesions against sham, whereas ANCHOR addressed predominantly classic lesions against verteporfin PDT.²³ End with a limitation: lesion classification and control standards reflect the study era, so their results should be read within protocol context. This is more accurate than listing a visual-acuity number from memory.

“What did CATT show?”

Begin with the factorial comparison: ranibizumab versus bevacizumab and monthly versus as-needed schedules in untreated neovascular AMD.⁴ State the broad first-year conclusion only if requested, then identify that as-needed treatment was protocolised and used fewer injections. Finally say that monitoring intensity, retreatment criteria and follow-up affect transferability. A short answer that includes the design will usually withstand a changed stem better than a memorised conclusion.

“Explain the VIEW regimen”

State that VIEW 1/2 tested aflibercept regimens against monthly ranibizumab, and that a key regimen was 2 mg every eight weeks after three monthly injections.⁶ The phrase “after three monthly injections” matters. Do not turn it into an assertion that every treatment course must follow it; it is a trial-protocol identifier.

Common acronym errors

Ten extra rapid-fire checks

21. What makes an OCT answer reproducible? Naming the compartment of fluid and scan limitation, not saying “wet”.
22. What was the ANCHOR comparator? Verteporfin PDT.³
23. What did MARINA use as its control? Sham injection.²
24. Why are exact percentages fragile recall? They vary by endpoint, arm and time point.
25. Does PRN necessarily mean fewer injections in every service? No; it depends on protocol and activity criteria.
26. Does treat-and-extend mean treatment stops once dry? No; the defining logic is proactive interval adjustment.
27. Which evidence source should settle a current drug-label question? The current regulator-approved label, not an old trial summary.
28. Which source should settle a current guideline question? The current professional guideline, not an original trial alone.¹
29. What does a trial’s primary endpoint tell you? The outcome it was designed and powered to assess, not every clinical consequence.
30. What is the core exam skill? Accurate study identification and bounded interpretation.

For topic-level revision, see the verified retina study guide. The Retina Deciphered page and free high-yield MCQs are existing OphthaMCQ resources. They are not AAO guidance and do not replace local protocols, regulatory information or specialist clinical judgement.

Sources

Footnotes

1. American Academy of Ophthalmology: Age-Related Macular Degeneration Preferred Practice Pattern. ↩ ↩²

2. MARINA: Ranibizumab for neovascular age-related macular degeneration. ↩ ↩² ↩³ ↩⁴

3. ANCHOR: Ranibizumab versus verteporfin for neovascular AMD. ↩ ↩² ↩³ ↩⁴

4. CATT: Ranibizumab and bevacizumab for neovascular AMD. ↩ ↩² ↩³ ↩⁴ ↩⁵

5. IVAN: Alternative treatments to inhibit VEGF in age-related choroidal neovascularisation. ↩ ↩²

6. VIEW 1 and VIEW 2: Intravitreal aflibercept for neovascular AMD. ↩ ↩² ↩³