Broad-spectrum sunscreens protecting from UV-radiation are useful in the management of melasma, but they failed to prevent relapses. The impact of the visible light on pigmentation is now demonstrated. However, using sunscreens effective against the entire visible light spectrum is almost impossible to use in daily practice. We recently demonstrated that short wavelengths of the visible spectrum (415 nm) can induce a prolonged hyperpigmentation in healthy volunteers while longer wavelengths (630 nm) do not affect pigmentation. Our aim was to evaluate the protective properties against melasma relapses of a sunscreen protecting against UVA/UVB and the shorter wavelengths of visible light compared to a sunscreen with UVA/UVB protection but without visible light protection.

We conducted a randomized controlled trial in real-life settings during the spring and summer (NCT 02061839). A photospectrometric analysis was first performed to assess the absorption potential of visible light of Formula A and Formula B. Both formulas contained the same filters against UV (octocrylene, methylene-bis-benzotriazolyltetramethylbutylphenol, butylmethoxydibenzoylmethanen bis-ethylexyloxyphenol, methoxyphenyl triazine, titanium dioxide). Formula A was tinted and contained iron oxides. The primary efficacy end point was the difference in the evolution of Melasma Area and Severity Index (MASI) score between the 2 groups. The calculated sample size was 40. Melasma patients were randomized to receive either Formula A or Formula B. They were trained on the right dose to be applied by using the revised teaspoon rule (1 teaspoon of sunscreen is applied to the face). Patients had to apply 1 dose of the allocated product twice daily and to apply an additional dose every 2 hours, 30 minutes before exposure to daylight. The main clinical assessment was the MASI score, performed on standardized pictures (VISIA, Canfield Scientific Inc) by 2 physicians masked to the type of sunscreen used. Table I (available at http://www.jaad.org) lists demographic and baseline data. One patient in group A was excluded because she also had postinflammatory hyperpigmentation. No patient was lost to follow-up.