Treating Acne With Topical Antibiotics

August 2019 | Volume 18 | Issue 8 | Editorials | 837 | Copyright © August 2019


Dr. Iain A. Stuart

Chief Scientific Officer Foamix Pharmaceuticals

Abstract
Letters to the Editor
Dear Editor,

I write to you to express concern about the recently published article by Bonati and Dover, Treating acne with topical antibiotics: current obstacles and the introduction of topical minocycline as a new treatment option.1 In articles of this type, it is fully understood that the intent behind presenting comparisons between established and/or emerging products is to further educate the dermatology community on pharmaceutical innovation occurring within the specialty. The article in question seeks to draw comparisons between two emerging topical products each containing minocycline at varying concentrations (BPX-01 and FMX101) and each being evaluated for the treatment of acne vulgaris. My primary concerns are specifically the following:

In Table 2, clinical study data is summarized from one Phase 2b study (N=225) evaluating BPX-01 and three Phase 3 studies (N=2,468) evaluating FMX101.3,4 What is absent from the article is that the data presented for BPX-01 2% for IGA treatment success was not statistically different from vehicle in this study. Equally, vehicle data from either clinical program has not been presented nor any statistical comparisons made at all. This is a key omission as the impact of a positive (or negative) vehicle effect is a well-established phenomenon in topical therapy irrespective of the disease under study and an important consideration in assessing the overall clinical utility of a product. Vehicle composition is as important at times as the pharmacologically active compound(s) as a recent article published in JDD outlines.5

Table 3 and corresponding narrative presents preclinical data evaluating systemic exposure of minocycline ofBPX-01 1 %, MNC-L 4%, and oral minocycline. Only BPX-01 2% clinical efficacy data is discussed in this article and therefore any inference linking low exposure with potential safety is meaningless from a clinical utility context. The use of data from a lipophilic, experimental formulation (MNC-L 4%) to draw inference on applicability to FMX101 4% is misleading, particularly when the composition of this experimental formulation was not disclosed.

Figure 3 provides a selective presentation of percent reduction of inflammatory lesion count reduction in clinical studies, again, excluding vehicle data and comparator statistics. Data from BPX- 01 1%, which was also evaluated in this study, is not included and was not statistically significantly different from vehicle at week 12.2 Specific week 2 data is not overtly presented in the plot and, again, was not statistically different from vehicle yet a statement of ">25% reduction" was included and the article concludes that BPX-01 2% is reported to have ... "a greater and quicker reduction than FMX101". The equivalent evaluation in the FMX101 clinical program was not completed at this timepoint. The authors go on to state for FMX101 that"... lesion counts began to separate after 3 weeks of treatment". This is inaccurate, the first post-baseline timepoint for this assessment was at week 3 and FMX101 was found to be statistically superior to vehicle. Data for BPX-01 at week 2 in the quoted Phase 2b study was not statistically different from vehicle. Data from FMX101 phase 3 program was not presented in equivalent plots in Figure 3 and demonstrates a lack of fair balance.

The statement that minocycline exposure as "undetectable" in 251 subjects from a study evaluating minocycline exposure when dosed as BPX-01 is a misrepresentation of the facts as the term should be "unquantifiable". The assay lower limit of quantification (LLOQ) is presented as 10 ng/ml, which is inappropriately high for dermal applications 5 and approximately 40-fold higher than the equivalent used to assess minocycline exposure when dosed as FMX101 (0.27 ng/ml).7 An LLOQ of 10 ng/ml is most definitely not a "highly sensitive assay" and has not been for a great many years. Reference 35 is miss-referenced in the article as it relates to oral minocycline product labelling and, as positioned, implies that oral minocycline was assessed in these studies. It was not.

Moreover, minocycline exposure when dosed orally was assessed in the corresponding study with FMX101 in a two-period, crossover study design and therefore a more meaningful comparison.7 No study information is provided in relation to the actual dose the subjects received, eg, grams per day, in these comparisons. The data presented for FMX101 was based on a maximum use safety study in subjects aged 18 years and older with a diagnosis of moderate-to-severe acne vulgaris using an inextremis dose of 4 grams per day for 21 days. The equivalent dosing used in the BPX-01 Phase 2b study is assumed to be 1 gram per day although not clear from the corresponding reference.

As it relates to the effect of ethanol on the disease state under study, ... the sequence of "ethanol...rapidly evaporates from the skin" "Ethanol also acts as an antimicrobial, exceeding the MIC and MBC for P acnes", " ... enter the pilosebaceous unit, where inflammatory and acne begin" is contradictory. If ethanol rapidly evaporates, one assumes that very little if any will enter the pilosebaceous unit to affect a meaningful antimicrobial action against C. acnes where, as the authors state, "inflammation and acne begin". The lack of vehicle clinical efficacy or safety information further clouds these statements. Although ethanol is referred to frequently within the article on its putative impact on acne vulgaris and beneficial effect on product formulation, at no point is there any discussion on the potential impact of the long-term use of primary alcohols on dermal integrity, eg, the potential for a deleterious effect on the stratum corneum and cutaneous adverse events.

Finally, like it's pathogenesis, the clinical presentation of acne vulgaris is multi-factorial although there is no discussion of either treatments respective impact on comedonal acne. Again, this speaks to a degree of selectivity in clinical data presentation as the effect on non-inflammatory lesion reduction in the