Comparison of narrowband ultraviolet B with psoralen plus ultraviolet A phototherapy for patients with early-stage mycosis fungoides: a systematic review and meta-analysis

Background The most prevalent cutaneous T-cell lymphoma is mycosis fungoides (MF), managed in early stages with psoralen plus ultraviolet A (PUVA) or narrowband ultraviolet B (NB-UVB) phototherapy. Objectives To evaluate the therapeutic effectiveness and side effect profiles of PUVA vs. NB-UVB in patients with early-stage MF. Method A systematic review was conducted, incorporating data from nine studies sourced from PubMed, Cochrane and Google Scholar. Eligible studies were those that simultaneously evaluated PUVA and NB-UVB, enrolling at least 10 adult participants per group with histologically confirmed early-stage (IA–IIA) MF and reported treatment outcomes. Exclusions were advanced disease, paediatric patients, noncomparative or nonrelevant treatments, small sample sizes or lack of outcome data. No language restrictions were applied. Key outcomes included any response, complete response, partial response, treatment failure, relapse-free interval and adverse effects. Results A total of 923 patients (age range 33–71 years; 52.5% men) were included: 556 treated with PUVA and 367 with NB-UVB. Any response was seen in 90.5% of those treated with PUVA vs. 88.3% of patients who received NB-UVB [odds ratio (OR) 1.18, 95% confidence interval (CI) 0.73–1.90; P = 0.50]. Complete response rates were 72% for those treated with PUVA and 61.8% for those treated with NB-UVB (OR 1.12, 95% CI 0.60–2.07; P = 0.73). Partial response was observed in 19.2% of patients who received PUVA and 28.0% of those who received NB-UVB (OR 0.87, 95% CI 0.45–1.69; P = 0.69). Treatment failure rates were lower with PUVA (9.7%) than with NB-UVB (12.6%) (OR 0.81, 95% CI 0.49–1.33; P = 0.41). Patients who received PUVA had a significantly longer median relapse-free interval (hazard ratio 1.94, 95% CI 1.07–3.51; P < 0.01). Adverse effects, including erythema, nausea, pruritus, burning, hyperpigmentation, pain, phototoxic reactions and polymorphic light eruption, showed no significant differences between groups. Conclusions Both PUVA and NB-UVB are effective, but PUVA may be preferred when sustained remission is the primary therapeutic goal.