LIGHTSITE III is the U.S. pivotal trial that supported the FDA’s De Novo authorization of the Valeda Light Delivery System. This page reports results as each source states them. The analyses are not identical, so we do not combine them into one unofficial number.

Study design (FDA and the papers agree on the outline)

LIGHTSITE III was a double-masked, randomized, sham-controlled, parallel-group, multicenter study in the United States. One hundred subjects (148 eyes) with dry AMD were randomized. The FDA decision summary states that 65 subjects (65%) received photobiomodulation (PBM) and 35 (35%) received sham, at 10 of 11 participating sites.

A treatment series was 9 sessions over 3 to 5 weeks, delivered every 4 months for 24 months (six series). Sham used a lower fluence of selected wavelengths; the 850 nm (near-infrared) light was omitted for sham, per the FDA summary and the papers. The primary effectiveness endpoint was the mean change in best-corrected visual acuity (BCVA) from baseline to Month 13 or Month 21 (Month 13 was tested first).

What the FDA authorization says (use this as the official claim)

The authorized indications include this sentence:

After about two years, the Valeda Light Delivery System treatment provides improved mean visual acuity of approximately one line of visual acuity (ETDRS) compared to those not receiving the treatment.

The FDA’s November 5, 2024 Roundup used similar wording: treatment after approximately two years “can provide an average improvement in vision equivalent to around one line on an eye chart.”

FDA decision-summary analysis (multiple imputation, mITT; PBM n = 91 eyes, sham n = 54)

From the DEN230083 decision summary:

• Month 13: the mean PBM result was 2.6 letters better than sham (p = 0.0548, not statistically significant at the protocol’s alpha of 0.025).
• Month 21 (the primary effectiveness timepoint the study met in this analysis): the mean PBM result was 3.8 letters better than sham (p = 0.0036; 95% confidence interval 1.2 to 6.3 letters).
• Month 24: PBM maintained a mean difference of 4.3 letters (95% CI 1.5 to 7.2).
• Within the PBM group, least-squares mean gains were 6.0 letters at Month 13, 6.2 letters at Month 21, and 5.6 letters at Month 24.
• There was no significant between-group difference at Months 13 and 21 in low-luminance BCVA or in Mars contrast sensitivity at 40 cm, and no significant between-group difference at Month 21 in macular drusen volume or thickness.
• Incident geographic atrophy, as counted in the summary: by Month 13, 5 of 50 sham eyes (10.0%) and 1 of 87 PBM eyes (1.1%); by Month 24, 12 of 50 sham eyes (24.0%) and 6 of 87 PBM eyes (6.8%). The FDA then cautions: “It is unclear if this outcome is related to the investigational treatment as development of new GA was not a pre-specified endpoint. The study and the pre-specified endpoints were not designed to assess whether the treatment could slow the rate of progression to GA, the incidence of GA, or the rate of growth of GA lesion.”
• Ocular adverse-event rates were similar between groups (about 25% of eyes). Procedure-related effects listed included punctate keratitis and an after-image in the sham group, and application-site warmth in the PBM group. One PBM subject had mild headache possibly related to the device.

13-month journal report (Boyer and colleagues, 2024)

Boyer D, Hu A, Warrow D, et al. LIGHTSITE III: 13-Month Efficacy and Safety Evaluation of Multiwavelength Photobiomodulation in Nonexudative (Dry) Age-Related Macular Degeneration Using the Lumithera Valeda Light Delivery System. Retina. 2024;44(3):487–497. doi:10.1097/IAE.0000000000003980. PubMed 37972955.

As stated in the PubMed abstract (retrieved November 15, 2026): the trial met the primary BCVA endpoint, with a between-group difference of 2.4 letters (SE 1.15; CI −4.7 to −0.1; P = 0.02) for PBM (n = 91 eyes) versus sham (n = 54 eyes). PBM alone: 5.4 letters. Sham alone: 3.0 letters. The PBM group showed a significant decrease in new-onset geographic atrophy (P = 0.024). A favorable safety profile was observed. These Month 13 figures are the authors’ reported analysis; they are not the same as the FDA’s Month 13 multiple-imputation p value above.

24-month journal report (Jaffe and colleagues, 2026)

Jaffe GJ, Boyer D, Hu A, et al. Long-term efficacy and safety of photobiomodulation in dry age-related macular degeneration (LIGHTSITE III: 24-month analysis). Retina. 2026;46(5):783–795. doi:10.1097/IAE.0000000000004822. PubMed 41791029.

As stated in the PubMed abstract (retrieved November 15, 2026): the study met the prespecified primary BCVA endpoint at Month 21 (P = 0.0036) with a +6.2 letter gain after PBM. At Month 21, 61.5% of PBM-treated eyes showed a gain of 5 or more letters, 23.1% gained 10 or more, and 4.4% gained 15 or more. A favorable safety profile was observed, with no signs of phototoxicity. Geographic atrophy incidence was 24.0% in sham eyes versus 6.8% in PBM eyes at Month 24 (P = 0.007). The authors also reported a significant benefit in vision quality of life. Read these figures alongside the FDA caveat on geographic atrophy above. The “+6.2 letter gain after PBM” matches the FDA within-group least-squares mean at Month 21; it is not the 3.8-letter between-group difference from the same FDA analysis.

What this does not show

LIGHTSITE III does not show that Valeda works for everyone, treats wet AMD, or is a cure. It does not, by FDA’s own comment, establish that the treatment slows geographic atrophy as a pre-specified claim. Duration of benefit after stopping treatment is uncertain.

Book an exam

You can choose your office and book an exam online, or call Hazleton Eye Specialists at 570-453-2020.

Medical disclaimer. This page is for general education only. It is not medical advice and does not replace an examination or a conversation with your own doctors. Do not start, stop, or change any medicine or treatment without talking to the clinician who cares for you. If you notice new wavy lines, a sudden blur, or a blank spot in your central vision, seek same-day care from an eye doctor or an emergency department.