Parametric Acne Severity (PAS) Score and Lesion Counts From Multi-Modality Facial Image Analysis Correlates Strongly with Investigator Assessment

June 2021 | Volume 20 | Issue 6 | Original Article | 642 | Copyright © June 2021


Published online May 14, 2021

Diane Thiboutot MD,a Amy Longenecker BSN,a Douglas Canfield BS,b and Sachin V. Patwardhan PhDb

aDepartment of Dermatology, The Pennsylvania State University College of Medicine, Hershey, PA
bCanfield Scientific, Inc., Parsippany-Troy Hills, NJ

Abstract
Variability in acne lesion counting and assessing global severity necessitates large sample sizes that increase trial costs. Lack of standardized measures for these outcomes precludes the conduct of meta-analyses needed to compare efficacy of acne treatments. The goal of this study was to evaluate objective measures of lesion counts and global severity using analysis of multimodal photography. An algorithm for counting lesions was trained and validated in 30 acne subjects and compared to parallel assessments by 2 expert raters. A composite of photographic data representative of acne lesion topography, erythema, and C Acnes fluorescence was used to generate a Parametric Acne Severity (PAS) score. No relationship was identified between lesion counts and IGA. The correlation coefficients between raters and the algorithm when compared per view for the inflammatory and non-inflammatory lesion counts were 0.77 (P=0.001) and 0.85 (P=0.001), respectively. The correlation coefficient between the raters’ IGA grades and the PAS score was 0.82 (P<0.05). These data demonstrate that the lesion counting, and PAS are objective measures that strongly correlate with investigator assessments. Inclusion of these measure in clinical trials may reduce variability, standardize outcomes, and provide insights into treatment effects on photographic parameters associated with acne.

J Drugs Dermatol. 2021;20(6):642-647. doi:10.36849/JDD.6165

INTRODUCTION

Acne is one of the most common skin conditions affecting millions of people worldwide. There are considerable negative effects of acne including lifelong scarring, feelings of low self-esteem, and an impaired quality of life.

Acne often requires treatment with combinations of drugs, many of which are expensive. Billions of US healthcare dollars are spent annually on the treatment of acne, both for office visits and for over the counter and prescription drugs, with little to no ability to estimate their comparative benefits. In this era of ever-changing paradigms in health care delivery, it is essential that health care providers base treatment decisions on data generated through evidence-based analyses.

The lack of standardization in how acne is assessed in clinical trials makes it difficult to pool data from different trials to assess comparable efficacy of different types of treatment. Health care providers need these data to guide treatment recommendations for patients with acne.

The Acne Core Outcomes Research Network (ACORN) was formed by an international group of experts with interest in improving how acne and its impacts are measured. The goal of the ACORN group is to develop a “toolbox” of validated measures to assess acne that can be adopted by researchers conducting clinical trials worldwide.1 The use of these standardized outcome measures would generate data that could be used to conduct meta-analyses to assess comparative to help guide treatment selection.

The primary end points of regulatory clinical trials for acne are acne lesion counts and the Investigator’s Global Assessment (IGA) of acne severity. However, both lesion counting and grading approaches have not been standardized. Furthermore, maintaining consistency within and across studies is challenging due to evaluators’ subjectivity. The Food and Drug Administration (FDA) recommends that study sponsors discuss their IGA scales and study designs with the FDA before trial implementation.2 Most IGA scales use phrases such as “none”, “few”, “several”, “moderate” or “many” inflammatory/non-inflammatory lesions as severity grade descriptors. Some scales also look at the lesion distribution using area of involvement description as “less than half”, “more than half”, or “entire area”. A study reported that dermatologists tend to