Evolution of Acne Assessments and Impact on Acne Medications: An Evolving, Imperfect Paradigm

January 2016 | Volume 15 | Issue 1 | Original Article | 79 | Copyright © January 2016


Linda Stein Gold MD,a Jerry Tan MD,b and Leon Kircik MDc

aHenry Ford Hospital, Detroit, MI
bUniversity of Western Ontario and Windsor Clinical Research Inc., Ontario, Canada
cMount Sinai Medical Center, NY, NY, Indiana University Medical Center, Indianapolis, IN; Medical Director, Physicians Skin Care, PLLC,
Louisville, KY; DermResearch, PLLC, Louisville, KY

Abstract
BACKGROUND: Outcomes for success in acne interventional trials include statistically significant differences from baseline between treatment arms in lesion counts (comedonal, inflammatory and/or total) and in thresholds of categorical improvement in investigator global assessments (IGA).
OBJECTIVES: We evaluated differences in outcome measures and definition of success in acne trials; and their impact on FDA approval and indications for acne medications.
METHODS: Review of acne clinical trial literature, prescribing information and regulatory guidelines for currently approved acne medications in the United States.
RESULTS: Numerous IGA scales exist with variations in specific categorical definitions. There are also differences in definitions of global success. Outcome success may not be accurately translated into corresponding terminology for indications.
CONCLUSIONS: Variability in IGA scales and definitions of success confound comparison of trial results for acne treatments. Harmonization and standardization of these factors will facilitate meta-analytics and treatment selection in patient care. Outcome measure success has not consistently been incorporated into acne medication indications.

J Drugs Dermatol. 2016;15(1):79-86.

INTRODUCTION

Progression of Acne Efficacy Assessments Over Time

In recent decades, there have been significant efforts across many medical disciplines to standardize outcome measures, thereby providing consistency and a means to pool results and compare treatment effectiveness.1 The objectives have been to determine relative benefits and harms with the intent of guiding clinicians and patients in selection of most appropriate interventions.2
Efforts to improve the rigor of classifications and outcome measures have also occurred in dermatology (Figure 1). Acne efficacy assessments have progressed through subjective clinical impression to today's stringent criteria of lesion counting plus global assessment of improvement plus a measure of treatment success (clear/almost clear and/or ≥2 grade improvement). 3 The increased rigor is based on understanding of the multifaceted aspects of acne and provides a greater spectrum of clinical measurement to more accurately reflect disease. However, it is important that these measures are evaluated for clinical relevance. Clinically irrelevant changes may be driven to statistical significance by large sample sizes, while use of inappropriate measures and definitions of success may relegate clinically relevant changes to non-significance statistically. In this article, we explore differences in outcome measures and definitions of success in acne trials; and their impact on United States Food and Drug Administration (FDA) approval and indications.

Acne Lesion Counting

Lesion counting was first reported in 1966 by Witkowski et al4 who determined the overall degree of improvement by the percentage decrease in lesions.4 Counting not only provides a more objective assessment compared with clinical impression5 but can also be divided into specific morphological types of lesions, allowing comparison of treatment effect by lesion type. Counts, being continuous variables, are sensitive to change (small changes can be detected) and are statistically powerful as a measure of change.
In 1996, Lucky et al studied acne lesion counting and reported good reliability when performed by the same rater but poor between raters.6 Additionally, the reliability of acne lesion counting decreased as the number of acne lesions increased.6 Subsequently, Tan et al showed better reliability in lesion counts among dermatologists, but somewhat less agreement in global assessments.7 Both studies reported that training improved results.6,7