We read with interest the CANDLE trial evaluating ibrexafungerp for prevention of recurrent vulvovaginal candidiasis. The study addresses a critical clinical need. We would, however, appreciate clarification regarding 2 methodological choices that are central to interpreting the results and are not discussed in the limitations section.
First, the protocol’s primary endpoint, “clinical success” at the test-of-cure visit, is defined as a test-of-cure evaluation with no mycologically proven, presumed, or suspected recurrence. In the Methods, mycologically proven recurrences require symptoms, positive microscopy, culture-confirmed Candida, and antifungal treatment; presumed recurrences require symptoms, positive microscopy, and antifungal treatment, but no culture confirmation; and suspected recurrences are defined solely by the use of antifungal therapy, including over-the-counter products, without requirements for symptom severity, microscopy, or culture. This means the primary endpoint merges culture-confirmed infection, symptom-based episodes, and treatment-only events. As a result, the outcome reflects not only the biology of recurrent vulvovaginal candidiasis and drug efficacy but also patient and clinician thresholds for empiric treatment and over-the-counter antifungal use. It would be valuable to know the proportion of failures attributable to each category and whether the treatment effect remains when limited to culture-confirmed recurrences.
Second, the handling of missing data for both clinical and mycologic outcomes applies a strict “missing=failure” rule. Participants who miss visits, withdraw early, or lack a culture at the relevant time point are automatically considered to have a recurrence or mycologic failure regardless of the reason for missingness. In a 36-week prophylaxis trial, absences may reflect adequate symptom control or loss to follow-up unrelated to disease status. If missingness differs by treatment arm, this assumption can bias effect estimates in either direction. The manuscript presents a sensitivity analysis excluding imputed outcomes but does not assess alternative approaches, such as time-to-event analyses or multiple imputation. Discussion of the impact of the “missing=failure” rule would strengthen confidence in the robustness of the findings.
Declaration of AI and AI-Assisted Technologies in the Writing Process
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
