Doxycycline: a first-line treatment for bullous pemphigoid?

www.thelancet.com Published online March 6, 2017 http://dx.doi.org/10.1016/S0140-6736(17)30549-4 1 Bullous pemphigoid is the most common autoimmune blistering skin disease, and incidence is on the rise, due at least in part to its association with older age. Treatment of bullous pemphigoid presents a challenge to the clinician, as first-line treatment regimens—either oral corticosteroids or whole body application of super-potent topical steroids—result in substantial morbidity and mortality or present logistical challenges in the elderly. Few trials have compared therapies for bullous pemphigoid. A previous landmark study found that whole-body topical application of the super-potent corticosteroid clobetasol propionate was at least as effective as oral prednisone for disease control at 3 weeks (and more effective for extensive disease). Notably, clobetasol proprionate treatment was associated with a significant reduction in mortality over 52 weeks compared with high-dose (1 mg/kg per day) but not medium-dose prednisone (0·5 mg/kg per day). Consistent with dose-dependent systemic absorption of topical steroids, reducing the dose of topical steroids provided a similar degree of disease control but with reduced morbidity and mortality. Tetracyclines have been used in dermatology for some time for their antiinflammatory properties. The only previous study comparing tetracyclines with corticosteroids in bullous pemphigoid was terminated early and the results, although promising, were not statistically significant. In The Lancet, Hywel Williams and colleagues report the results of the largest multicentre trial comparing oral treatments for bullous pemphigoid; the investigators and participating sites should be congratulated for completing this valuable study. Using a non-inferiority trial design, they tested the hypothesis that 200 mg/day oral doxycycline is not inferior in effectiveness to oral prednisolone (0·5 mg/kg per day) and that doxycycline is less likely to result in severe adverse effects. Participants were allowed to apply up to 30 g per week of potent topical corticosteroids to localised lesions for the first 3 weeks, and after 6 weeks. A high proportion of participants were available for analysis at 6 weeks although, perhaps not unexpectedly, a substantial number dropped out by week 52. Rates of withdrawal were balanced between the two treatment groups. So what were the major findings? The modified intention-to-treat analysis showed that disease control at 6 weeks was achieved by 91% (92 of 101) of participants starting prednisolone compared with 74% (83 of 112) starting doxycycline—an adjusted difference (by baseline severity of bullous pemphigoid and Karnofsky score) of 18·6% (90% CI 11·1–26·1). In terms of safety, this trial underscored the adverse event profile of systemic steroids for bullous pemphigoid: 36% (41 of 113) of participants starting medium-dose oral prednisolone developed a serious adverse event over 52 weeks compared with 18% (22 of 121) of participants starting doxycycline—an adjusted difference of 19·0% (95% CI 7·9–30·1). In agreement with this, and previous studies, treatment-related deaths were greater in the prednisolone group (11 deaths) compared with the doxycycline group (three deaths). So how should these data be interpreted? Some help is provided within the protocol and statistical analysis plan and by an earlier publication from the group that considers, in advance, possible outcomes from this trial. Additionally, it is important to consider how the trial was powered and how the non-inferiority margins were determined and set. It appears that the trial was primarily powered to detect a 20% difference in grade 3, 4, and 5 side-effects. Achieving a consensus around the setting of the non-inferiority margin is recognised to be challenging. The authors commendably addressed this in part by undertaking a survey of dermatologists Doxycycline: a first-line treatment for bullous pemphigoid?