eTable 1. Morphea Subtypes and Definitions Designated by Previously Proposed Classification Schemes
eTable 2. Demographic and Clinical Characteristics of Morphea by Age at Onset
eTable 3. Patient-Reported Features and Comorbidities by Age of Onset
Customize your JAMA Network experience by selecting one or more topics from the list below.
Identify all potential conflicts of interest that might be relevant to your comment.
Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.
Err on the side of full disclosure.
If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.
Not all submitted comments are published. Please see our commenting policy for details.
Prasad S, Zhu JL, Schollaert-Fitch K, Torok KS, Jacobe HT. An Evaluation of the Performance of Current Morphea Subtype Classifications. JAMA Dermatol. 2021;157(4):399–405. doi:10.1001/jamadermatol.2020.5809
How frequently do the commonly used morphea classification systems successfully categorize patients with clinically relevant morphea subtypes?
In this cross-sectional study that included 944 adult and pediatric patients from 2 prospective cohorts, the Padua criteria successfully classified 95% of patients in comparison with the Peterson criteria, which classified 56% of patients, and the European Dermatology Forum classification, which classified 52% of patients.
The Padua criteria performed best in classifying patients into groups with cohesive demographic and clinical features, but they haveambiguities that need to be addressed by consensus-based approaches.
Numerous classification systems for morphea subtypes exist, but none have been systematically evaluated for their ability to categorize patients with morphea into demographically and clinically coherent groups. Although some subtypes, such as linear morphea, are present across all the classification schemes, others list unique subtypes. This creates confusion among investigators and practitioners and impairs accurate categorization of patients for study and clinical evaluation.
To evaluate how frequently the commonly used morphea classification systems categorize patients with morphea into clinically relevant subtypes using cross-sectional analysis of 2 large patient cohorts.
Design, Setting, and Participants
This cross-sectional study comprised 944 adults and children from 2 prospective cohorts—the Morphea in Adults and Children at the University of Texas Southwestern Medical Center (Dallas, Texas), which enrolled participants from July 20, 2007, to September 21, 2018, and the National Registry for Childhood-Onset Scleroderma at the University of Pittsburgh (Pittsburgh, Pennsylvania), which enrolled participants from October 23, 2002, to November 13, 2018.
Main Outcomes and Measures
Patient demographic characteristics, morphea subtype, quality-of-life measures, disease activity, and damage as measured by Localized Scleroderma Cutaneous Assessment Tool scores during initial visits.
A total of 944 participants (444 patients with adult-onset morphea and 500 patients with pediatric-onset morphea; 741 female participants [78%]; median age at onset, 16 years [interquartile range, 8-44 years]) were included in this study. Most participants were White (723 [77%]) and had the linear (474 [50%]) or generalized subtype of morphea (244 [26%]). With the use of the previously published Padua criteria, most patients were classified to have linear morphea (474 [50%]), followed by generalized morphea (244 [26%]), plaque morphea (141 [15%]), mixed morphea (38 [4%]), and pansclerotic morphea (3 [0.3%]). Overall, the Padua criteria successfully classified 900 patients (95%) in comparison with the Peterson criteria (533 [56%]) and the European Dermatology Forum classification (487 [52%]).
Conclusions and Relevance
In this cross-sectional study of morphea subtype classification systems, the Padua criteria performed best in classifying patients into subgroups with cohesive demographic and clinical features, supporting its widespread use. However, they have ambiguities that might lead to misclassification, particularly in terms of generalized and pansclerotic morphea and descriptors such as morphea profunda. Consensus-based approaches are needed to address these ambiguities and develop a unified classification scheme.
Morphea, or localized scleroderma, is an inflammatory condition of the skin and soft tissue that results in excessive collagen deposition, often producing permanent functional and cosmetic impairment. Morphea affects both adults and children, with approximately 50% of patients having childhood-onset disease.1,2 To date, to our knowledge, few studies have been sufficiently powered to provide in-depth analysis of the relevant clinical features of morphea disease subtypes across the life span. Currently published work is predominantly retrospective, evaluates relatively small cohorts, includes only adult or pediatric patients, or is focused exclusively on rheumatology or dermatology.3-5 As a result, the demographic and clinical features of morphea, particularly of the less-frequent subtypes, remain poorly characterized, and little is known about the differences between adults and children with morphea.
Consequently, numerous classification systems for morphea subtypes are currently in use (eTable 1 in the Supplement). In 1995, Peterson et al6 published the Mayo Clinic Criteria, which classified morphea into the 5 following subtypes: plaque, generalized, bullous, linear, and deep. In contrast, Laxer and Zulian7 published the Padua criteria in 2006, which outlined the following 5 different subtypes: circumscribed (superficial and deep), generalized, linear (trunk or limbs and head), pansclerotic, and mixed, but the Padua criteria do not include bullous morphea or deep morphea. However, they do note deep involvement as a modifier, which can occur with circumscribed lesions, but not as a type on its own. In 2017, the European Dermatology Forum proposed a classification system with the following 6 subtypes: (1) limited, which includes plaque, guttate, and superficial morphea; (2) generalized, which includes generalized morphea and pansclerotic subtypes; (3) linear, which includes linear morphea of extremities, en coup de sabre, and Parry-Romberg syndrome; (4) deep; (5) mixed; and (6) eosinophilic fasciitis.8 Although all of these classification systems are used in practice, they have limitations. To our knowledge, there are no data on how well these different classification systems perform in accurately categorizing patients with morphea into clinically and demographically coherent groups. This negatively impacts patient care as well as multisite collaborations for both observational and interventional studies in which it is important to know that patients are categorized consistently in terms of determining associated disease outcomes.
The Morphea in Adults and Children (MAC) cohort3 and the National Registry for Childhood-Onset Scleroderma (NRCOS)5 prospectively collect clinical, demographic, and quality-of-life features of adults and children with morphea. Together, they include more than 900 patients with morphea. The objective of the present study was to examine the MAC and NRCOS cohorts in a cross-sectional manner to evaluate which classification schemes characterize the greatest proportion of patients into cohesive, clinically relevant subtypes and to identify areas of ambiguity for further refinement.
This study included a total of 944 participants with morphea enrolled in the prospective MAC and NRCOS cohorts. The criteria for inclusion included eligibility for enrollment in the MAC or NRCOS cohorts (the details of eligibility and characteristics of these cohorts have been previously reported).3,5 The MAC cohort included 441 adults and 262 children enrolled between July 20, 2007, and September 21, 2018, and the NRCOS cohort included 3 adults and 238 children enrolled between October 23, 2002, and November 13, 2018. All participants who had subtype classification data available from the initial visit were included in the study. All participants or legal guardians of child participants provided written consent for inclusion in their respective cohorts, which was approved by the University of Texas Southwestern Medical Center or the University of Pittsburgh institutional review board.
Demographic, clinical, and quality-of-life data were collected on each patient at his or her initial study visit. One physician (K.S.T. or H.T.J.) examined all of the participants at their respective institution using predetermined case report forms. Because the case report forms used in both cohorts contain morphea subtype classifications, such as linear, circumscribed, and the others that encompass existing classification systems, participants were assigned a subtype based on existing major classification schemes.7 Deep, dermal, and superficial modifiers were used only in the MAC case report forms. All lesions were scored using the Localized Scleroderma Cutaneous Assessment Tool (LoSCAT), a validated outcome measure that includes components of disease activity (modified Localized Scleroderma Skin Severity Index [mLoSSI]) and damage (Localized Scleroderma Skin Damage Index [LoSDI]).9,10 Accompanying the LoSCAT, the Physician Global Assessment of Disease Activity (PGA-A) and Physician Global Assessment of Disease Damage (PGA-D) were scored for all patients. The raters (K.S.T. and H.T.J.) were involved in the development and validation of the LoSCAT and its components to be included in the physician global assessments, ensuring good agreement between raters.9 Quality-of-life information included the Skindex-29 + 3,11 the Dermatology Life Quality Index (DLQI) for adults,12 the Children’s DLQI (cDLQI),13 and patient-reported symptoms, such as pain and itch, on a 10-point visual analog scale. Additional clinical information on cutaneous and extracutaneous manifestations were included, such as depth of involvement (superficial, dermal, or deep) and functional limitations (limitation of range of motion, contractures, and limb-length discrepancy). Superficial involvement was defined as lesions with histologic features of lichen sclerosus et atrophicus. Dermal involvement was defined as lesions that affected the dermis from the papillary to the reticular. Deep involvement was defined as subcutaneous involvement and was evaluated via imaging and/or histologic testing.
To ensure consistency across the 2 raters (dermatologist [H.T.J.] and rheumatologist [K.S.T.]), a group of 21 patients with morphea were recruited and examined simultaneously (20 minutes between raters in a blinded fashion) at the University of Texas Southwestern Medical Center as a separate substudy to evaluate interrater reliability and agreement of the LoSCAT components (mLoSSI, LoSDI, PGA-A, and PGA-D).9 The raters also reviewed the different subtype classifications and the clinical features that constitute each to ensure consistent subtyping. Interclass correlation coefficients were calculated to quantitatively evaluate agreement between raters.
Demographic and clinical variables were reported as either frequency counts or median values with interquartile ranges (IQRs). Mann-Whitney tests were used to compare clinical characteristics composed of continuous data between onset category (pediatric vs adult). The χ2 test and the Fisher exact test were used to compare dichotomous or categorical variables. A 2-sided P < .05 was considered statistically significant. All statistical analyses were performed with IBM SPSS Statistics for Windows, version 25.0 (IBM Corp).
For the training session in which both raters assessed 21 patients with morphea, the interclass correlation coefficient between raters was 0.92 for mLoSSI, 0.98 for LoSDI, 0.83 for PGA-A, and 0.82 for PGA-D. Interclass correlation coefficient between raters for morphea subtype was 0.89. Patients for whom there was disagreement between subtype were reassessed by raters at the time of the meeting, and ambiguities were resolved.
Table 1 displays the clinical and demographic characteristics of all participants. Of the total 944 participants, 500 (53%) had pediatric-onset morphea, and 444 (47%) had adult-onset morphea. Most patients (723 [77%]) were White. The median age at onset was 16 years (IQR, 8-44 years) overall, 9 years (IQR, 5-12 years) for the pediatric-onset group, and 45 years (IQR, 31-58 years) for the adult-onset group. Although female patients comprised the majority of both pediatric-onset and adult-onset patient groups, there was a higher proportion of male patients in the pediatric-onset group (26% [132 of 500]) than in the adult-onset group (16% [71 of 444]).
With the use of the published Padua criteria, 95% (900 of 944) of patients could be categorized (Table 2).6-8 Most patients were classified as having linear morphea (474 [50%]), followed by generalized morphea (244 [26%]), plaque morphea (141 [15%]), mixed morphea (38 [4%]), and pansclerotic morphea (3 [0.3%]). We were unable to classify a subtype for 44 patients (4.7%) using the Padua criteria; these patients were classified as having indeterminate morphea. Of these patients, 3 were classified as having eosinophilic fasciitis. The patients with indeterminate morphea were predominantly female (86% [38 of 44]), and White (68% [30 of 44]) and were more likely to have adult-onset morphea (68% [30 of 44]). When these patients were evaluated at subsequent follow-up visits, 2 major patterns emerged—plaques that eventually evolved to meet criteria for generalized subtypes and plaques of the neck and trunk in children that evolved to the linear subtype. Furthermore, in the MAC cohort, patients who met criteria for the pansclerotic subtype were concomitantly classified as having generalized morphea because these patients simultaneously met criteria for both (n = 16).
In contrast, the Mayo Clinic classification criteria categorized 56% (533 of 944) of patients (Table 2).6-8 Morphea profunda, a subtype according to this classification criteria, or deep lesions were observed in 52% of patients (367 of 703) and occurred across different demographic groups and lesional morphologic characteristics. Among those classified as having morphea profunda by the Mayo classifications criteria, 65% (224 of 342) had lesions in a linear distribution, 34% (76 of 222) had lesions in a generalized distribution, and 47% (45 of 95) had plaque or circumscribed lesions. No patients were classified as having keloidal or bullous morphea, subtypes that are proposed in the Mayo classification criteria. Bullae were reported in dependent sites, such as the lower extremity overlying morphea lesions, and keloids were recorded in patients with skin of color with a predisposition toward keloids, as evidenced by keloids in unaffected sites.
The European Dermatology Forum classification criteria categorized 52% of patients (487 of 944) (Table 2).6-8 Superficial morphea, which includes epidermal and superficial dermal involvement, was observed in 17% of patients (116 of 703). Similar to morphea profunda, superficial morphea occurred across all subtypes—10% (34 of 342) had linear distribution, 29% (64 of 222) had generalized distribution, and 12% (11 of 95) had plaque distribution.
With the use of the Padua criteria, most patients whose morphea was classified as linear had pediatric-onset disease (74% [349 of 474]), with a median age at onset of 11 years (IQR, 6-18 years). In contrast, adult-onset disease was more common in patients with generalized morphea (82% [201 of 244]). Compared with patients with linear morphea, those with generalized morphea had higher median mLoSSI (13 [IQR, 5-27] vs 3 [IQR, 0-6]; P < .001), LoSDI (19 [IQR, 12-28] vs 8 [IQR, 5-14]; P < .001), and PGA-A (30 [IQR, 15-60] vs 15 [IQR, 0-40]; P < .001) scores (Table 1). In contrast, PGA-D scores were higher among patients with linear morphea than among those with generalized morphea (30 [IQR, 20-50] vs 20 [IQR, 12-40]; P < .001). When comparing superficial and deeper involvement by subtype, we found that patients with generalized morphea were more likely to have superficial involvement compared with patients with linear morphea (29% [64 of 222] vs 10% [34 of 342]) and were less likely to have deep involvement compared with patients with the linear subtype (34% [76 of 222] vs 65% [224 of 342]).
Patients with the linear subtype were also more likely than patients with generalized morphea to have joint deformity (8% [40 of 474] vs 0.4% [1 of 244]) and limited range of motion (26% [123 of 474] vs 20% [50 of 244]) (Table 1). Patients with generalized morphea were more likely than those with linear morphea to have functional abnormalities in their upper extremities (14% [34 of 244] vs 9% [43 of 474]). The frequency of lower extremity functional abnormalities was similar between patients with linear morphea and patients with generalized morphea (17% [80 of 474] and 16% [40 of 244], respectively).
Similar trends were seen when comparing patients by age at onset. Children were more likely than adults to have joint deformity (11% [53 of 500] vs 1% [4 of 444]; P < .001) and limited range of motion (25% [124 of 500] vs 18% [78 of 444]; P = .008) (eTable 2 in the Supplement). Patients with pediatric-onset morphea were more likely than those with adult-onset morphea to have functional abnormalities in the trunk (ie, back and shoulder) (4% [19 of 500] vs 2% [10 of 444]; P = .02) and lower extremity (18% [88 of 500] vs 13% [59 of 444]; P = .003) (eTable 2 in the Supplement).
Quality of life, as measured by the DLQI and cDLQI, was most commonly categorized as mild for patients with linear morphea (55% [263 of 474]) and those with generalized morphea (48% [116 of 244]). Table 3 lists patient-reported symptoms and comorbidities. Joint pain was more frequent for patients with generalized morphea than for those with linear morphea (36% [88 of 244] vs 25% [117 of 474]; P = .008). Among comorbidities, patients with generalized morphea reported higher rates of depressive symptoms than patients with linear morphea (5% [12 of 244] vs 2% [9 of 474]). Psoriasis was more common in patients with generalized morphea than in those with linear morphea (4% [9 of 244] vs 1% [5 of 474]). Juvenile idiopathic arthritis (in children) and rheumatoid arthritis (adult onset) were more common in patients with linear morphea than in those with generalized morphea (2% [10 of 474] vs 1% [3 of 244]) (eTable 3 in the Supplement).
There is substantial ambiguity surrounding morphea subtypes, which has resulted in the use of different classification schemes by different groups of investigators, making it difficult to categorize patients for research and clinical care. In the present cross-sectional study, we examined a large prospective cohort of patients with adult-onset or pediatric-onset morphea to evaluate the proportion of patients whose morphea was categorized into a specific subtype using different morphea classification systems. We also evaluated whether the proposed subtypes would help us to identify clinically and demographically cohesive groups.
Our results demonstrated that the Padua criteria capture the most relevant disease subtypes in morphea. Using the Padua criteria, we were able to categorize 95% of patients (900 of 944) as having specific subtypes of morphea. In comparison, the Mayo Clinic (Peterson et al6) criteria were able to categorize 56% of patients (533 of 944), and the European Dermatology Forum criteria were able to categorize 52% of patients (487 of 944). Furthermore, using the Padua criteria to categorize patients, we found that the identified subtypes consisted of cohesive clinical and demographic features, particularly in terms of patients with linear lesions having pediatric onset and less female predominance, whereas generalized morphea was most prevalent in postmenopausal female patients with adult-onset disease.1,14-16
In contrast to the Padua criteria, other classification schemes categorized patients into subtypes that were not present in this cohort or that were not associated with specific demographic and clinical features or morphologic characteristics. The Peterson et al6 criteria designated deep morphea as its own subtype termed morphea profunda, defined by “more diffuse involvement of the deep dermis, subcutaneous tissue, fascia, or superficial muscle that did not demonstrate the linear pattern.”6(p1070) We found that the presence of deep morphea occurred in children and adults and in those with lesions in a linear, generalized, and plaque or circumscribed distribution, which mirrors prior observations.14,17-19 Similarly, when examining the features of bullous or keloidal morphea, as described in the Peterson et al6 criteria, we found that bullous lesions occurred exclusively in dependent areas of the extremities that had deep, circumferential plaques in those with linear and generalized lesion distributions. Keloidal lesions were present in patients with skin of color overlying plaques of morphea and outside areas affected with morphea with linear or generalized morphologic characteristics. Because deep, bullous, and keloidal lesions occurred across different groups defined by demographic features and lesion morphologic characteristics, these terms may be better served as modifiers or descriptors rather than as subtypes.
The European Dermatology Forum classification scheme, which includes deep morphea as a subtype (defined by “fibrotic process mainly affecting the deeper layers typically arranged symmetrically on the extremities”),8(p1404) also designated superficial morphea as a subset of limited morphea (defined as “symmetrical, single or multiple, sharply demarcated, hyperpigmented, nonindurated patches located on the trunk or extremities”).8(p1404) We observed that the presence of superficial morphea occurred across ages and lesion morphologic characteristics and was particularly pronounced in postmenopausal patients with generalized morphea, similar to a previously published study.17 No patients were classified as having guttate morphea, a subtype designated by this classification scheme.
Despite the strong performance of the Padua criteria, our results indicate that clarification is warranted (Box).7 Specifically, the Padua criteria define generalized lesions as “induration of the skin starting as individual plaques (≥4 and larger than 3 cm) that become confluent and involve at least 2 out of 7 anatomic sites.”7 However, using this definition, patients with multiple linear lesions would also meet the criteria for generalized disease. As has been reported previously, because generalized morphea is so distinct from linear morphea both clinically and demographically, patients with multiple linear lesions should be categorized as having linear morphea even when they technically meet criteria for generalized lesions based on the body sites affected.17 Similarly, in the Padua criteria, pansclerotic is as “circumferential involvement of limbs affecting the skin, subcutaneous tissue, muscle and bone.”7 However, this distribution may be seen in patients with deep lesions affecting multiple body sites who meet the criteria for generalized morphea. Patients with multiple linear lesions with deep involvement may also meet the criteria for pansclerotic morphea, which indicates that clarification of the pansclerotic subtype is needed. None of the current classification schemes address issues surrounding linear lesions of the head, hemifacial atrophy, or extracutaneous manifestations, such as inflammatory arthropathy and neurologic findings. Furthermore, the inclusion of entities such as eosinophilic fasciitis, which coexists with typical morphea lesions in up to 30% of cases, has not been addressed, to our knowledge.20 In contrast, the coexistence of systemic sclerosis and morphea lesions was not observed. These issues will also need to be considered in future classifications so that these important entities maybe more accurately studied.
Trunk or limbs
This study has some limitations, including that both cohorts were derived from tertiary care centers, which likely led to overrepresentation of the more severe morphea subtypes and presentations. Furthermore, more adults are enrolled in the MAC cohort, limiting their evaluation to 1 rater. This was mitigated by coordinating the raters’ assessment of subtype, clinical features, and scoring via an in-person session.
Taken together, the results of our study have several implications for practice and future multisite collaborative studies, encompassing both dermatology and rheumatology. First, the Padua criteria had the highest frequency of classifying patients with similar morphologic characteristics of cutaneous lesions and demographic characteristics, highlighting its promise as a classification scheme for morphea for both clinical and research purposes. However, the Padua criteria would benefit from further refinement to resolve ambiguities that may lead to misclassification. Specifically, we found that the definition of the generalized subtype and extracutaneous involvement merit further examination. We propose that subtypes (such as deep and superficial) that are present in other classification schemes and are helpful in further classifying the distribution and depth of involvement be used as modifiers in the existing Padua criteria. Consensus-based approaches that include multidisciplinary, international stakeholders are needed to ensure uptake of a unified classification scheme for morphea that addresses the disease across the life span.
Accepted for Publication: December 25, 2020.
Published Online: February 17, 2021. doi:10.1001/jamadermatol.2020.5809
Corresponding Authors: Heidi T. Jacobe, MD, MSCS, Department of Dermatology, University of Texas Southwestern Medical Center, 5323 Harry Hines Blvd, Dallas, TX 75390 (firstname.lastname@example.org); Kathryn S. Torok, MD, Pediatric Rheumatology, Children’s Hospital of Pittsburgh, 4401 Penn Ave, Pittsburgh, PA 15224 (email@example.com).
Author Contributions: Dr Prasad and Ms Zhu had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Dr Prasad and Ms Zhu contributed equally to this work and are co–first authors. Drs Torok and Jacobe are co–senior authors.
Concept and design: Prasad, Zhu, Torok, Jacobe.
Acquisition, analysis, or interpretation of data: Prasad, Zhu, Schollaert-Fitch, Torok.
Drafting of the manuscript: Prasad, Zhu, Schollaert-Fitch, Torok.
Critical revision of the manuscript for important intellectual content: Zhu, Torok, Jacobe.
Statistical analysis: Prasad, Zhu, Schollaert-Fitch.
Obtained funding: Jacobe.
Administrative, technical, or material support: Schollaert-Fitch, Jacobe.
Supervision: Torok, Jacobe.
Conflict of Interest Disclosures: Dr Torok reported receiving grants from Scleroderma Foundation, Nancy Taylor Foundation for Chronic Diseases, and Scleroderma Research Foundation outside the submitted work. No other disclosures were reported.
Funding/Support: Research reported in this publication was supported in part by grant K23AR056303-5 from the National Institutes of Health (Dr Jacobe), grant K23 AR059722 from the National Institutes of Health (Dr Torok), and grant RO1 R01AR073516-01A1 from the National Institutes of Health (Drs Jacobe and Torok); by the National Center for Advancing Translational Sciences of the National Institutes of Health under award TL1TR001104; with support from UT-STAR, NIH/NCRR/NCATS grant UL1TR000451; and the Nancy Taylor Foundation for Chronic Diseases Inc.
Role of the Funder/Sponsor: The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Disclaimer: The content is solely the responsibility of the authors and does not necessarily represent the official views of UT-STAR, The University of Texas Southwestern Medical Center at Dallas and its affiliated academic and health care centers, the National Center for Research Resources, or the National Institutes of Health.