Early effective treatment and reliable review matter; a fixed six- or three-injection rule and one universal extension schedule do not fit every drug or patient. Non-inferiority and observational improvement do not prove superiority or cure.
Terminology and draft status
Terminology review is not complete. Selected actual WHO UMD queries and their limits are recorded in terminology.md; they do not approve the whole translation.
01Separate centre involvement from the need to inject
Determine the cause, OCT morphology, best-corrected acuity and foveal involvement before treating macular oedema. The original 2019 German pathway is updated by the 2025 S3 guideline: non-centre-involving sight-threatening clinically significant DME may be offered selected focal laser, with attention to retinal damage; it is not an automatic first-line requirement for every noncentral oedema. 1 2
With foveal DME and relevant visual impairment, offer anti-VEGF when morphology supports a prospect of visual benefit. With centre involvement but no relevant visual impairment, the 2025 guideline instead supports observation without immediate IVOM or laser, with controlled reassessment. Centre involvement alone is not an instruction to inject; examine the whole retina and manage systemic diabetes and vascular risk as well. 1 2
For insufficient response, recheck diagnosis, ischaemia, traction, structural damage, actual treatment exposure and follow-up before changing drugs. A switch to another VEGF inhibitor or selected steroid may be considered, but OCT drying alone does not guarantee visual recovery. Steroid selection requires pressure, cataract and implant-related risk assessment, not simply fewer appointments. 2
02Individualise loading and ongoing review
The original recommends six monthly anti-VEGF injections for DME, a check one month later and another one to three monthly doses; for RVO it gives three initial monthly injections. Retain these as the cited teaching pathway, not a single regimen for all products or eyes. Current loading and retreatment depend on the drug, regional label, vision/OCT response and cause; do not delay an indicated review until a numerical series is complete. 1 2 11
PRN separates monitoring from activity-triggered injections; the original monthly-review description is protocol-dependent. T&E gives an active injection at the treatment visit and adjusts the next interval according to activity, often in two- or four-week steps. It can be used in unilateral or bilateral disease with coordination; it is not principally restricted to having only one affected eye. Dose interval and total service visits are different burdens. 1 2 11
Explain why undertreatment, access problems and missed visits can worsen outcomes without blaming patients. The original Freiburg 44% insufficient adherence in year 1, Moorfields more-than-half unable to relax, one-quarter sleep disturbance and 42% wanting fewer injections lack the exact cohorts, questions and definitions here. Anxiety, carers, travel, diabetes appointments and insufficient clinic capacity still warrant practical support; these figures are not universal patient probabilities. 1
03Compare durability with drug-specific safety and trial design
KESTREL/KITE compared brolucizumab with aflibercept 2 mg and met the four-letter non-inferiority margin for vision at 52 weeks, with different loading schedules and selected longer brolucizumab intervals. Better anatomical drying does not prove superior vision. Inflammation, retinal vasculitis and vascular occlusion are important risks; assess for inflammation before treatment, do not inject with active inflammation, and investigate new symptoms urgently. No superiority over all flexible alternatives follows. 1 2 7
YOSEMITE/RHINE randomised 1891 adults with centre-involving DME to faricimab every 8 weeks, an individualised 4–16-week faricimab schedule or fixed aflibercept 2 mg every 8 weeks. The one-year endpoint averaged weeks 48/52/56 and met a four-letter non-inferiority margin. This supports trial-specific extension, not superior vision or fewer visits compared with every flexible aflibercept T&E strategy; the comparator was not allowed equivalent extension. 1 2 6
Intraocular infection, inflammation, pressure elevation and retinal complications remain relevant with all injection plans. A reassuring trial safety profile cannot exclude rare postmarketing events. Product-specific contraindications and the current local information must be checked; frequency reduction is not a safety guarantee or permission to omit monitoring. 2 5
04Interpret PHOTON using its population and denominators
PHOTON was a randomised double-masked phase 2/3 trial in adults with centre-involving DME.660 were randomised and 658 treated: 8q12 n=328, 8q16 n=163, 2q8 n=167. The original calls them treatment-naive, but the accessed US study description states 44% had prior DME treatment. 8 mg arms started with 3 monthly doses,2 mg with 5; shortening from week 16 and selected extension from week 52 followed predefined criteria. 1 3 5
The primary 48-week non-inferiority margin was 4 letters. Mean observed gains 8.8/7.9/9.2 letters differ from adjusted-model comparisons: 8q12 versus 2q8 −0.57 (95% CI−2.26 to 1.13), 8q16 −1.44 (−3.27 to 0.39). Both met the prespecified margin, not superior vision. Randomisation/masking support the comparison, while manufacturer funding, selected eligibility and unread full analysis details limit generalisation. 3
By 96 weeks 534 completed follow-up. The published abstract reports mean vision gains 8.8/7.5/8.4 letters and mean active injections 9.5/7.8/13.8 for 8q12 / 8q16 / 2q8. The original six fewer injections,7.8 versus 13.8, is the group difference over this period, not a guaranteed personal saving. All arms reduced thickness; anatomy and functional benefit are different endpoints and the extended arms did not have identical loading exposure. 1 4
Among 96-week completers,88% in 8q12 and 83% in 8q16 maintained their respective≥12/≥16-week categories;43%/47% qualified for≥20 weeks. The original “nearly 90%≥12 weeks” and pooled 44%≥20 weeks need those definitions; last assignment differs from an interval maintained throughout. Similar ocular event rates,37.1–45.4% across arms in the abstract, and no new signal do not exclude rare inflammation or pressure risk; a numerically lower IOP rate does not prove protection. 1 4 5
05Treat venous oedema and continue neovascular surveillance
Distinguish central, branch and hemiretinal occlusion. CRVO often has more extensive ischaemic/neovascular risk; BRVO can also threaten central vision. The original “almost half develop severe complications” lacks a defined population and endpoint and must not become every patient’s risk. Treat visually significant oedema promptly and examine for iris/angle/retinal neovascularisation separately; a dry OCT does not remove that need. 1 11
CRYSTAL enrolled 357 patients with CRVO in a24-month prospective open-label single-arm study. Minimum 3 monthly ranibizumab injections continued until stability criteria; up to month 23 the median was 15 and mean 13.1 injections. This supports explaining potentially prolonged treatment but does not make 15 a universal two-year requirement or prove one regimen superior without a comparator. 1 8
The original 60% requiring≥30 injections over 5 years and persistent vision loss after a treatment interruption require their exact cohorts and interruption definitions. Retinal thickness may improve after restarting while visual function does not fully recover; neither all interruptions cause irreversible loss nor all anatomical recoveries restore vision. Identify and address follow-up barriers early. 1
T&E can support organised RVO care and sustained monitoring; its aim is disease control with a feasible burden, not necessarily the fewest injections. Drug choice, ischaemia, function and neovascular risk guide interval changes. Continue tailored surveillance when treatment is paused or stopped, especially in ischaemic CRVO. 1 11
06Interpret switching and an exit case without promising cure
The original registry account gives 527 patients, +10 letters,44% gaining 15 letters and 19% switching, with drug-specific 11%/43%/38% switch rates. The identified 36-month FRB report concerns 527 treatment-naive CRVO EYES, +10 letters (95% CI 7–12) overall and 30% switching;257/527 (49%) completed. Its denominators and timing do not reproduce all original percentages, which remain unresolved rather than silently reassigned. Selection and missing follow-up limit whole-cohort outcome estimates. 1 9
A switch may reduce thickness without adding significant vision. Historical availability, starting drug, baseline severity and why a switch was chosen can explain different switching proportions; a lower rate is not proof of drug superiority. The original later-market aflibercept explanation is plausible context, not a tested causal explanation. Assess the individual eye before changing therapy. 1 9
AURIGA was a prospective descriptive observational study, not a randomised switch comparison. In 554 naive and 65 previously treated RVO patients,24-month mean gains were 11.4 letters (95% CI 9.4–13.3) and 4.4 (−0.6 to 9.5), respectively. This supports possible routine-care improvement, but the previously treated interval includes no change and cannot prove every switched CRVO benefits or that persistence matters more than drug choice. Confounding, attrition and baseline differences remain important. 1 10
The lecture proposes better six-month thickness reduction and acuity as response markers and describes early venous-stasis oedema treated until a16-week interval, then stopped without recurrence in the reported follow-up. These are associations and an uncontrolled teaching case with no defined long-term observation here, not validated cure criteria. RVO oedema can resolve, but ongoing vascular risk and recurrence monitoring remain; choose any exit individually rather than by reaching 16 weeks alone. 1
What are the main learning points?
Early effective treatment and reliable review matter; a fixed six- or three-injection rule and one universal extension schedule do not fit every drug or patient. Non-inferiority and observational improvement do not prove superiority or cure.
When is assessment urgent?
New visual deterioration or distortion requires prompt assessment. Severe pain, redness, light sensitivity or marked loss after an injection needs immediate eye-emergency review. Do not skip, stop or change treatment without a plan agreed with your team.
References and access status
- 1Immutable original export; DME/RVO therapy update; case and survey provenance unverified
Complete export read; original attribution and any recording/transcript or appointment not independently verified.
- 2German S3 diabetic retinal complications 2025, Version 1.0
Selected DMEsection 2.2.6 printedpp 16–22 and opening recommendation 2-15 read 13 November 2026; not whole 81 pages/method report/every cited study. Do not copy the internally suspect visual-acuity MD 48.10 in steroid discussion as verified letters.
- 3PHOTON 48 week phase 2/3 trial
Complete NCBI abstract read 13 November 2026; full report, protocol, tables and supplements not appraised.
- 4PHOTON 96 week report 2026
Complete NCBI abstract read 13 November 2026; full report, protocol, tables and supplements not appraised.
- 5EyleaHD US prescribing information, April 2026
ActualPDF selected 2.2/2.3/2.5 and PHOTON 14.2 study-description text read 13 November 2026, including 44% previously treated. Not entire Table 4/figures or German/EU prescribing validation; injection versus follow-up visit distinguished.
- 6YOSEMITE/RHINE faricimab phase 3 one-year report
Complete NCBI abstract read 14 November 2026; full report, protocol, tables and supplements not appraised.
- 7KESTREL/KITE brolucizumab 52 week report
Complete NCBI abstract read 14 November 2026; full report, protocol, tables and supplements not appraised.
- 8CRYSTAL 24 month single-arm study
Complete NCBI abstract read 14 November 2026; full report, protocol, tables and supplements not appraised.
- 9FRB registry CRVO 36 month outcomes
Complete NCBI abstract read 14 November 2026; full report, protocol, tables and supplements not appraised.
- 10AURIGA observational RVO 24 month outcomes
Complete NCBI abstract read 14 November 2026; full report, protocol, tables and supplements not appraised.
- 11RCOphth 2022 RVO guideline executive summary
Selected cached primaryPDF printedpp 2–5 reread 13 November 2026;2022 guideline, not a new 2026 evidence base from directory date. Full guideline/figures and every study not appraised.
For patients
Macular swelling can have different causes. Some eyes need frequent injections and others can extend or be observed with planned checks. Tell the team about anxiety, travel or appointment problems so that treatment remains feasible.
Examination and selected tests guide individual treatment and follow-up; this educational text cannot replace your own assessment.
New visual deterioration or distortion requires prompt assessment. Severe pain, redness, light sensitivity or marked loss after an injection needs immediate eye-emergency review. Do not skip, stop or change treatment without a plan agreed with your team.
New visual deterioration or distortion requires prompt assessment. Severe pain, redness, light sensitivity or marked loss after an injection needs immediate eye-emergency review. Do not skip, stop or change treatment without a plan agreed with your team.