Rosacea is an inflammatory skin condition that affects approximately 5% of the adult population, with the highest prevalence in Europe and North America.1 Despite its prevalence, rosacea remains poorly understood from a pathophysiologic perspective, with no diagnostic laboratory markers.2 Because diagnosis relies on clinical judgment, the nomenclature for describing and characterizing rosacea becomes paramount in ensuring that patients are given an accurate diagnosis and subsequent treatment. We review the shortfalls in the recent history of rosacea classification and discuss their implications.
Subtype to Phenotype Classification
In 2002, the National Rosacea Society (NRS) Expert Committee published a standardized classification schema for rosacea (Table).3 The authors described primary and secondary diagnostic criteria. The presence of 1 or more primary features in a central facial distribution was indicative of rosacea. Primary characteristics included flushing (transient erythema), nontransient erythema, papules and pustules, and telangiectasia. Secondary features, which could occur with or independently of primary features, included burning or stinging of the face, dry appearance, facial edema, ocular manifestations, peripheral (nonfacial) occurrence, phymatous changes, and red facial plaques. Whereas these features often present simultaneously in a characteristic pattern, they were grouped into 4 main subtypes—erythematotelangiectatic (ETR), papulopustular, phymatous, and ocular—and 1 variant, granulomatous rosacea.3
To enhance clinical and research applications of this categorization system as well as offer further standardization, the NRS released a supplementary clinical grading scorecard in 2004 in which each of the primary and secondary characteristics could be assigned a subjective severity score of absent, mild, moderate, or severe. The goal was that the subtype classification and clinical grading system, when used in conjunction with each other, would establish a common language for patients, clinicians, and researchers to describe and further investigate rosacea.4
The 2002 categorization system was certainly an impactful first step in the organization of rosacea. It was not without its critics, however, namely rosacea-oriented dermatologists who were concerned about its lack of specificity.5-7 For instance, the NRS Expert Committee did not address the time frame for flushing, which typically has a long duration in rosacea patients, or for the nontransient erythema; telangiectasia secondary to heliodermatitis; or the often-observed periocular sparing. Additionally, the schema did not account for conditions such as gram-negative folliculitis (pustules characteristically located on the central face) or discuss the need to rule out carcinoid, mastocytosis, or connective-tissue disease, which can lead to nontransient facial erythema. Without strict definitions and exclusions, nonrosacea disorders could be incorrectly labeled as rosacea.
Beyond the lack of specificity, there was additional concern if a subtype system was the ideal way to capture disease presentation and severity. By subtyping, there was unnecessary division of interrelated disease into individual disorders; an individual’s clinical presentation might fall along a spectrum rather than within a discrete box.8
Furthermore, from a research standpoint, subtyping rosacea could hinder or confuse epidemiologic studies. For instance, if patients present with phenotypes from different subtypes, into which subtype would they fall?8-10
The global ROSacea COnsensus (ROSCO) panel, comprising 17 international dermatologists and ophthalmologists, convened in 2016 to address this matter. The panel proposed a new system (published in 2017) based on individual phenotypes.9 In this new system, diagnostic features include persistent centrofacial erythema with periods of increased intensity and phymatous changes. Major features, which are diagnostic when there are at least 2, include flushing (transient erythema), inflammatory papules and pustules, centrofacial telangiectasia, and ocular manifestations. Each feature could then be graded on a severity spectrum independent of concurrent phenotypes (Table).8
The panel concluded that this system would provide a stronger foundation for standardization as new knowledge of rosacea continues to be elucidated.8 In support of their argument, ROSCO also released a treatment algorithm that depended on a phenotype scheme.11 The panel emphasized that by focusing on individual lesions rather than a subtype encompassing many characteristics, treatment could be tailored to the patient. Using this à-la-carte therapy option, physicians could choose those rosacea aspects that are particularly concerning to the patient and manage only those aspects or overlap treatments to improve multiple aspects.11
In 2017, 15 years after the original classification system was proposed, the NRS updated their classification system (published in 2018), taking into consideration some of the criticisms as well as new scientific data on rosacea. Similar to the schema proposed by ROSCO, this system was based on phenotype. Inclusion and exclusion criteria were more robust in this update compared to the original classification in 2002. The criteria provide a timeline for transient flushing—it must occur within seconds or minutes in response to a neurovascular stimulant—and state that it is characteristically prolonged (Table).12