Outdated Frameworks, Missed Opportunities: A Call for Establishing Classification Criteria for Patients with Sarcoidosis

Abstract

Progress in sarcoidosis research is hampered by inconsistent definitions and study entry criteria that rely on variable local practice patterns and widely differing expert opinions. This, in turn, has resulted in a heterogeneous body of clinical trial data that often cannot be compared, replicated, or generalized. The rheumatology community has successfully distinguished between diagnostic and classification criteria for diseases such as systemic lupus erythematosus, IgG4-related disease, and systemic sclerosis. This task was also successfully accomplished for patients diagnosed with pulmonary diseases, such as pulmonary arterial hypertension, asthma, and chronic obstructive pulmonary disease, for which classification has established distinct treatment pathways between groups. Diagnostic criteria are used at the bedside, on a patient-by-patient basis. They reflect the needs of health care providers and clinicians to consider disease heterogeneity and exercise judgment. Classification criteria, however, are primarily used for research and aim for specificity to define homogeneous populations for study. Unfortunately, sarcoidosis does not have definite classification criteria. Their absence undermines proper interpretation of study outcomes and hampers development of novel therapies. Sarcoidosis can be thought of as “constellations of patient populations” that need to be unified. As clinician-researchers and, above all else, advocates for our patients, we must move beyond diagnosis of exclusion and define sarcoidosis by what it is rather than by not what it is not.

Supplementary Information

The online version contains supplementary material available at 10.1007/s40744-026-00838-1.

Key Summary Points

Classification criteria used in clinical research are frequently confused with diagnostic criteria, which are employed at the bedside.

Lack of sarcoidosis classification criteria hampers clinical trial execution and outcomes, as well as development of new therapies.

Classification criteria have already been published for systemic immune diseases such as rheumatoid arthritis, systemic lupus erythematosus, IgG4-related disease, and systemic sclerosis, among others.

Sarcoidosis classification criteria are being developed by clinician-scientists at several sarcoidosis centers of excellence, led by the University of Illinois Chicago.

Unmet Medical Need in Sarcoidosis

Sarcoidosis remains a frustrating paradox for clinicians and researchers alike. It is simultaneously well-known and poorly defined. Rheumatologists, with their backgrounds in multisystem diseases, are often involved in the care of patients with sarcoidosis, many of whom suffer from multisystem involvement, unpredictable disease trajectories, and limited therapeutic options [1]. Despite decades of research and evolving therapeutic strategies, the field still lacks validated disease classification criteria. This absence has hindered clinical trials, delayed therapeutic innovation, and contributed to fragmented care across specialties. Classification criteria are not merely academic tools. In rheumatology, robust criteria for diseases such as systemic lupus erythematosus (SLE) (Table 1) [2], IgG4-related disease, systemic sclerosis, and others have galvanized research, clarified phenotypes, and enabled high-quality clinical trials.

Table 1.

Example of rheumatology classification criteria: main structure of the EULAR/ACR 2019 disease classification criteria for systemic lupus erythematosus

Obligatory entry criterion: ever positive ANA (≥1:80 or equivalent)
≥10 points of 10 domains (highest item of each domain counted only)
1. Renal	Class III/IV LN 10	Class II/V LN 8	Proteinuria 4
2. Musculoskeletal	Joint involvement 6	–	–
3. Serosal	Acute pericarditis 6	Effusion 5	–
4. Mucocutaneous	ACLE 6	SCLE or discoid LE 4	Alopecia or oral ulcers 2
5. Neuropsychiatric	Seizure 5	Psychosis 3	Delirium 2
6. Hematologic	Autoimmune Hemolysis 4	Thrombocytopenia 4	Leukopenia 3
7. Constitutional	Fever 2	–	–
8. Specific antibodies	Anti-dsDNA 6	Anti-Sm 6	–
9. Low complements	C3 and C4 low 4	C3 and C4 low 3	–
10. APS antibodies	LAC 2	ACLA or anti-β2gpI 2	–

EULAR/ACR European Alliance of Associations for Rheumatology/American College of Rheumatology, ANA anti-nuclear antibodies, LE lupus erythematosus, ACLE acute cutaneous LE, SCLE subacute cutaneous LE, LN lupus nephritis, dsDNA double- stranded DNA, Sm, Smith antigen, APS anti-phospholipid syndrome, LAC lupus anticoagulant, ACLA anti-cardiolipin antibodies, anti-β2gpI anti-β2-glycoprotein I antibodies

[2] Reprinted from Aringer M, et al. Best Pract Res Clin Rheumatol. 2023;37:101,949, an open-access article distributed under the terms of the Creative Commons CC-BY license

By contrast, sarcoidosis research has been hampered by inconsistent definitions and study entry criteria that rely on variable local practice patterns or expert opinion. The result is a heterogeneous body of clinical trial data that often cannot be compared, replicated, or generalized. Sarcoidosis is a systemic granulomatous disease with protean manifestations. Organ involvement spans systems from pulmonary, ocular, cardiac, neurologic, cutaneous, and others. While excellent organ-specific diagnostic frameworks exist—such as the International Workshop on Ocular Sarcoidosis (IWOS) criteria for ocular sarcoidosis or the 2014 World Association of Sarcoidosis and Other Granulomatous Disorders (WASOG) instrument for organ involvement—none of these serve as global classification criteria for the disease. Most have not been validated for use in patient selection for clinical trials, and none address the fundamental question researchers face (i.e., How can we reliably identify a patient as having sarcoidosis for the purpose of research?).

Diagnostic criteria are used at the bedside. They reflect the clinicians’ needs to encompass disease heterogeneity and exercise judgment. Classification criteria, by contrast, aim for specificity, providing a shared language to define homogeneous populations for study. Sarcoidosis sorely needs such a language.

A major barrier to the creation of classification criteria for sarcoidosis has been the legacy dependence on biopsy-proven non-caseating granulomas as the cornerstone of diagnosis. While histologic confirmation remains important, it is often not feasible or necessary across all organ systems. Moreover, many granulomatous mimickers—such as tuberculosis, berylliosis, and certain malignancies—confound interpretation. A well-developed classification system should integrate clinical, radiographic, histologic, and immunologic domains, weighing them appropriately to reflect their discriminatory values. This has been successfully achieved in other systemic diseases using multi-criteria decision analysis (MCDA) and expert consensus methodologies.

In 2025, the rheumatology team at the University of Illinois Chicago (UIC) launched a multi-phase initiative to address this gap. Beginning with a comprehensive literature review, which is captured here, and to be followed by a Delphi consensus survey involving international experts from WASOG and the Foundation for Sarcoidosis Research (FSR) Global Sarcoidosis Clinic Alliance Members, the team is constructing a derivation cohort to empirically define and weight candidate items. Ultimately, the goal is to develop and validate a sarcoidosis classification system with high specificity for research use—just as rheumatology has done successfully for other diseases during the past decade.

As sarcoidosis treatment moves beyond corticosteroids toward biologics and precision immunomodulation, the need for standardized research entry criteria becomes increasingly more urgent [3].

The distinction between classification and diagnostic criteria has been successfully implemented for patients diagnosed with other pulmonary diseases, such as but not limited to pulmonary arterial hypertension, asthma, and chronic obstructive pulmonary disease (COPD), for which classification grouped similar patients by disease severity. In these cases, this algorithmic approach enabled clinicians and investigators to deploy consistent treatment pathways in these diagnostically diverse patient populations. For example, the Global Initiative for Chronic Obstructive Lung Disease has created, published, and regularly updated the GOLD classification criteria (created in 1997 and first reported in 2001), which covers the full spectrum of COPD and its heterogeneous patient populations (Fig. 1) [4].

Fig. 1.

Classification of chronic obstructive pulmonary disease. LABA long-acting beta agonist, LAMA long-acting muscarinic antagonist, ICS inhaled corticosteroids, eos eosinophils, mMRC modified Medical Research Council, CAT COPD Assessment Test. [4] ©2026, Global Initiative for Chronic Obstructive Lung Disease, available from www.goldcopd.org, published in Deer Park, IL, USA. Commercial reproduction not permitted

We invite our colleagues across disciplines—rheumatology, pulmonology, ophthalmology, cardiology, dermatology, and neurology—to join us to create the foundational tools for improved clinical trial endpoints that our patients deserve. It is time to move beyond diagnosis of exclusion. Let us define sarcoidosis not by what it is not, but by what it is.

This article is based on previously conducted studies and does not include any new studies with human participants or animals performed by any of the authors.

Classification Versus Diagnostic Criteria in Sarcoidosis

Disease classification criteria are developed primarily for research, to ensure that patients enrolled in studies share the same underlying condition. They prioritize specificity over sensitivity, which reduces heterogeneity in study populations, but may exclude borderline or atypical cases. In contrast, diagnostic criteria are intended for clinical practice, for which sensitivity is often more important, to avoid missing patients who need treatment. Although classification criteria are perhaps being used in daily clinical care—for example, in SLE, systemic sclerosis, antiphospholipid syndrome, and IgG4-related disease—patients who do not fulfill them may still be considered by treating physicians as likely having these diseases, but should not be included.

The main differences between diagnostic and classification criteria [5] are summarized in Table 2.

Table 2.

Differentiating diagnostic criteria from classification criteria

Diagnostic criteria

Diagnostic criteria are generally broad and must reflect the different features of a disease (heterogeneity), with a view to accurately identifying as many people with the condition as possible.

Given this complexity, development and validation of diagnostic criteria can be quite challenging.

Few validated diagnostic criteria exist in rheumatology, and clinicians commonly establish a diagnosis based on subjective combination of clinical signs/symptoms, available clinical tests, and knowledge about the epidemiology of their geographical areas.

Classification criteria

Classification criteria are standardized definitions primarily intended to create well-defined, relatively homogenous cohorts for clinical research. They are not intended to capture the entire universe of possible patients, but rather to capture the majority of patients with key shared features of the condition.

Goals of classification differ from the intent of diagnostic criteria. Validated classification criteria are considered critical to the interpretation of study findings and comparisons of results between studies.

Classification criteria traditionally have high specificity, which generally comes at the expense of somewhat lower sensitivity. Consequently, few individuals are incorrectly labeled as having a disease (“false positives”), but a percentage of individuals with the disease diagnosis may be “missed.” This may make classification criteria inappropriate for use in routine clinical care.

Sarcoidosis and Classification Criteria

Sarcoidosis is a multisystem disease characterized by granuloma formation in one or more organs. The clinical phenotype is highly variable, depending on the affected organs, and highly unpredictable [1, 6, 7]. Among both pulmonologists and rheumatologists alike, sarcoidosis is now considered a systemic disease more than a pulmonary one [1]. Despite significant advances in imaging technology, no criteria or guidelines exist on using modern imaging techniques, including high-resolution computer tomography, in sarcoidosis disease classification and phenotyping [8–11].

Treatment has evolved from the almost exclusive use of corticosteroids, first noted to be efficacious in the early 1950s [12, 13], to the use of non-FDA-approved medications such as disease-modifying antirheumatic drugs (DMARDs) including traditional synthetic DMARDs (e.g., methotrexate, azathioprine, mycophenolate mofetil, leflunomide), biologic DMARDs (e.g., TNF inhibitors, rituximab) and targeted synthetic DMARDs (e.g., Janus kinase inhibitors). Corticosteroids are approved only for pulmonary sarcoidosis [14].

Literature Review of Existing Sarcoidosis Criteria

To determine whether diagnostic or classification criteria have been developed for sarcoidosis, we performed a scoping literature review of classification and diagnostic criteria for sarcoidosis. We evaluated the medical literature in English for criteria that either classify or diagnose systemic sarcoidosis, ocular sarcoidosis, cardiac sarcoidosis, and neurosarcoidosis. The methodology for this scoping review, including the literature databases searched (e.g., Medline/PubMed and Embase), search terms, eligibility criteria, screening process, and a PRISMA flow diagram are provided in a supplementary appendix to this article.

After screening 2981 abstracts, we found 22 relevant articles [15–36]. Six articles proposed diagnostic criteria for sarcoidosis in general [15, 19, 21, 25, 27, 28], eight for cardiac sarcoidosis [16, 20, 22, 26, 29–32], three for ocular sarcoidosis [17, 23, 24], and four for neurosarcoidosis [18, 33–35]. Of the 22 articles reviewed, only one [36] proposed classification criteria, and it was limited to ocular involvement.

Limitations of Existing Sarcoidosis Criteria and Instruments

ATS/ERS/WASOG Statement

The most important and most frequently cited criteria set for the diagnosis of sarcoidosis is the American Thoracic Society (ATS)/European Respiratory Society (ERS)/WASOG Statement on Sarcoidosis, published in 1999 [27], which states, “the diagnosis of sarcoidosis needs a compatible clinical picture, histologic demonstration of non-caseating granulomas, and exclusion of other diseases capable of producing a similar histologic or clinical picture. The presence of non-caseating granulomas in a single organ such as skin does not establish a diagnosis of sarcoidosis.” This statement reflected a consensus of experts who provided guidance to clinicians on making a diagnosis of sarcoidosis and not for classifying patients for research purposes. Although this work was performed by excellent clinicians with extensive expertise in sarcoidosis, to our knowledge, these criteria have never been validated, and the detailed consensus methodology was not described in detail.

ACCESS Instrument

In 1999, Judson et al. published the initial version of “A Case Control Etiologic Study of Sarcoidosis (ACCESS) Proposed Instrument” [28]. The goal was to “develop a comprehensive sarcoidosis assessment instrument to standardize the definition of pulmonary and extrapulmonary involvement” and was used to “assess organ involvement in ACCESS patients” who were followed up for 2 years. This instrument was developed by “a consensus of opinions of the members of the steering committee of ACCESS based on review of their experience and medical literature” [28]. The method behind the development of these criteria was not otherwise described. This instrument proposed criteria to determine organ involvement in patients with biopsy-confirmed sarcoidosis with three possible degrees of confidence: “definite,” “probable,” or “possible” [28]. This statement was followed in 2007 by a single-author opinion article by Judson, who suggested proposals for diagnostic criteria [25].

WASOG Organ Assessment Instrument

Drawing on data from the ACCESS cohort [28], in 2014, WASOG published a new sarcoidosis organ assessment instrument [21]. Using a multidisciplinary panel and a Delphi exercise, the group developed an instrument that “may be useful for the clinician and researcher in establishing criteria for sarcoidosis organ involvement.” Experts developed possible criteria for individual organ involvement, which were then presented to a group of experts for evaluation via a Delphi exercise. These criteria rely on presentation of various clinical conditions “in the setting of histologic evidence of granulomatous inflammation of unknown cause, demonstrated in at least one organ that is not being assessed, and that the clinical manifestation and granulomatous inflammation not be due to any other alternative causes” [21]. In addition, the clinical manifestation being assessed required that “all alternative causes other than sarcoidosis for this clinical manifestation had been reasonably excluded.” For example, for eye disease, the presence of uveitis would be considered highly probable for eye sarcoidosis if a biopsy from another organ indicated typical granuloma and another cause of that had been ruled out. Although widely used and relied upon in confirmation of sarcoidosis diagnoses, the WASOG organ assessment instrument has yet to be validated.

Sarcoidosis Diagnostic Score (SDS)

In 2018, Bickett et al. [19] proposed a Sarcoidosis Diagnostic Score (SDS) derived from the WASOG organ involvement criteria. The proposed calculation instrument can be used for either patients with positive biopsy (SDS Biopsy) or patients with a clinical diagnosis (SDS Clinical). This prospective study employed data from patients at the University of Cincinnati Medical Center Sarcoidosis and Interstitial Lung Disease Clinic collected over 6 months. One physician made a diagnosis of sarcoidosis based on clinical features consistent with sarcoidosis and ≥ 1 biopsy demonstrating non-caseating granuloma with no alternative diagnosis identified. Mimickers were patients with non-sarcoidosis who presented with signs or symptoms consistent with sarcoidosis, but for whom an alternative diagnosis (e.g., granulomas of known causes, local sarcoid reactions, berylliosis, and immune deficiency syndrome) was confirmed [19].

An initial cohort was used to derive the scoring system. A separate group of patients served as the validation cohort. Each organ involved was classified by one physician as biopsy positive, highly probable, at least probable, or possible based on the WASOG criteria. The goal was to assign a numeric score to each of these categories and then, for each patient, to sum the weights for all involved organs and create a total score. Sensitivity and specificity for the total score versus the physician’s diagnosis of sarcoidosis vs. mimickers were calculated and receiver operating curves (ROCs) were used to determine cutoffs. With weights for each organ of 3 for “highly probable” and 2 for “at least probable” for both SDS Clinical and SDS Biopsy, they found a total patient score ≥ 6 for SDS Biopsy and ≥ 3 for SDS Clinical yielded good sensitivity and specificity [19]. The SDS was validated in a cohort of 380 patients, including 103 with sarcoidosis [19].

This paper most closely resembles current classification criteria in that consensus was used to develop the individual organ criteria, weighting was assigned, and data were used to develop cutoff scores and to validate the criteria. Moreover, it resembles classification criteria by prioritizing specificity and comparison with a set of mimickers. However, weaknesses of the methodology, which we will address in our project, include that only one physician’s diagnosis of sarcoidosis was used as the “gold standard” and that the physicians were not asked to decide if they would enter the case in a research study, an important aspect of classification versus diagnostic criteria. The weighting of each organ’s degree of involvement was done by expert opinion and not empirically tested. Classification criteria prioritize specificity to produce homogeneous research cohorts, acknowledging a trade-off in sensitivity. To prove specificity, comparison with a set of mimickers is needed. Culver and colleagues recently published a framework for listing such mimickers [37].

The detailed methodology of item generation and reduction for the WASOG organ criteria have not been published. The SDS has not been used as an entry criterion in any studies, but rather in a meeting abstract that reported validation in a large international cohort [38]. It has also been recently validated in a multicontinental study (Table 3) [39].

Table 3.

Summary of Sarcoidosis Diagnostic Score Clinical and Sarcoidosis Diagnostic Score Biopsy to differentiate patients with sarcoidosis from control patients without sarcoidosis

	Score criteria	Sensitivity (%)	Specificity (%)	LR+	LR−
SDS Clinical	≥3	93.7	72.2	3.4	0.1
	≥5	67.5	88.3	5.8	0.4
	≥10	10.4	99.8	53.7	0.9
SDS Biopsy	≥5	100.0	85.8	7.0	0.00
	≥6	98.5	92.9	14.0	0.02
	≥12	41.2	99.0	42.6	0.6

LR1 positive likelihood ratio, LR2 negative likelihood ratio, SDS Sarcoidosis Diagnostic Score

Reprinted with permission of the American Thoracic Society

Copyright © 2026 American Thoracic Society. All rights reserved

[39] Jeny F, et al. Ann Am Thorac Soc. 2023;20:371–80

Annals of the American Thoracic Society is an official journal of the American Thoracic Society

American Thoracic Society Diagnostic Criteria

A 2020 ATS statement [15] focused on specific aspects of diagnosis using population, intervention, comparison, outcome (PICO) methodology but did not make specific PICO recommendations regarding sarcoidosis diagnostic criteria. The stated objectives of that paper were to “make recommendations that address uncertainties that are commonly confronted by clinicians relating to the diagnosis and detection of sarcoidosis.” A multidisciplinary panel of experts in sarcoidosis constructed clinically important questions related to diagnostic testing for sarcoidosis and the panel used the Grading of Recommendations, Assessment, Development, and Evaluation (GRADE) approach to formulate and grade the strength of the recommendations. The panel concluded, “The diagnosis of sarcoidosis is not standardized, but is based on three major criteria: a compatible clinical presentation, the finding of non-caseating granulomatous inflammation in one or more tissue samples (not always required), and the exclusion of alternative causes of granulomatous disease.” As noted by the authors, “Presently, there are no established objective measures to determine if each of these diagnostic criteria has been satisfied, and, therefore, the diagnosis of sarcoidosis is never fully secure” [15].

Classification of Sarcoidosis by Organ Involvement

In the diagnostic criteria for cardiac sarcoidosis [22], there are “two pathways to a diagnosis: 1) histologic diagnosis from myocardial tissue: Cardiac sarcoidosis is diagnosed in the presence of non-caseating granulomas on histologic examination of myocardial tissue with no alternative cause identified (including negative microbiology testing if applicable); and 2) clinical diagnosis from invasive and non-invasive studies. Cardiac sarcoidosis is probable if a) histologic diagnosis of extra-cardiac sarcoidosis is present; b) one or more of following is present (see cardiac sarcoidosis criteria); and c) other causes for the cardiac manifestations have been reasonably excluded.”

Similarly, diagnostic criteria for ocular sarcoidosis (OS) were initially published in 2009 [23] and were revised in 2019 [17]. These diagnostic criteria for ocular sarcoidosis categorized “definite OS,” “presumed OS,” and “probable OS.” In 2021, the Standardization of Uveitis Nomenclature (SUN) Working Group (SUN) published the classification criteria specifically for sarcoidosis-associated uveitis [36]. The SUN group found that, “The key features of the criteria were a compatible uveitic syndrome and evidence of sarcoidosis” [36]. Adalimumab remains the only biologic approved by the FDA for the treatment of non-infectious uveitis [40].

In 2018, Stern et al. [18] published work to “propose diagnostic criteria for central nervous system and peripheral nervous system neurosarcoidosis.” Like the other diagnostic criteria, the neurosarcoidosis diagnostic criteria developed three possibilities: “possible” when there is no pathologic confirmation; “probable” when the pathology reveals granulomas; and “definite” when the pathology “is consistent with neurosarcoidosis.” Although some of these organ-specific criteria, especially the ocular criteria, were developed with excellent methodology, it is not clear that these patients would meet internationally agreed upon classification criteria for sarcoidosis. That would be important if the findings, for example, of biologic studies can be extrapolated to sarcoidosis in general. On the other hand, if this works as well as the other organ involvement work described above, it will be especially useful in the process of the development of new classification criteria for sarcoidosis in general.

GRADS Study Classification Framework

The GRADS study enrolled patients with sarcoidosis based on established ATS/ERS diagnostic criteria [27], requiring either histologic confirmation of non-caseating granulomas or a compatible clinical presentation (e.g., Löfgren syndrome) after exclusion of alternative diagnoses [41]. Importantly, GRADS was designed as a deep-phenotyping and translational study rather than a diagnostic validation effort, and enrollment criteria were intentionally broad to capture biological heterogeneity. Nevertheless, like other currently available diagnostic or classification frameworks, the GRADS approach has inherent limitations. Diagnostic certainty relied on investigator consensus and standard clinical workup, rather than a uniform reference standard. In addition, biopsy yield remained organ-dependent and imperfect, and extrapulmonary involvement, particularly cardiac sarcoidosis, was not systematically screened at enrollment. Moreover, exclusion of active smokers and patients unable to undergo bronchoscopy may have introduced selection bias, limiting generalizability.

Summary of Current Sarcoidosis Criteria and Statements

Of the six papers on the criteria for sarcoidosis itself [15, 19, 21, 25, 27, 28], none generated an item list solely from the literature—two [21, 27] relied on an expert panel, and two [15, 28] used a combination of literature review and expert opinion. Of the four papers that used expert opinion, only one [21] conducted a formal item-reduction process.

Sensitivity and specificity of the criteria were assessed in cohorts of patients and mimickers by only one study [32]. None of the articles defined in detail the items in the criteria.

The WASOG/ACCESS criteria are neither classification nor diagnostic criteria for sarcoidosis. They are criteria to define organ involvement in patients with an established diagnosis of sarcoidosis. The subsequent work of Bickett et al. [19] did validate the WASOG organ involvement criteria, demonstrating some face validity because the criteria were based on expert opinion via a Delphi and used patient data. A degree of construct validity was found, as the criteria were associated with a positive physician diagnosis of sarcoidosis. This information will be important in helping us develop items for the new classification criteria.

In addition to the diagnostic criteria for systemic sarcoidosis itself, diagnostic criteria are published for isolated organ involvement. Ideally, these would be criteria to decide if specific organ involvement is a part of sarcoidosis in a patient already diagnosed by criteria for sarcoidosis. However, in most cases, these criteria start with the organ involved in patients not yet diagnosed with sarcoidosis. This is the case for ocular sarcoidosis [17], cardiac sarcoidosis [22], and neurosarcoidosis [28].

Discussion

Bridging the Gap in Unmet Need

On the basis of a review of ClinicalTrials.gov conducted in 2024, 305 studies in sarcoidosis are underway. Of the 305 studies, 182 are interventional and 122 are observational. As evidenced by the preliminary results of our scoping review, no study currently recruiting patients with sarcoidosis does so using classification criteria with known sensitivity and specificity. Lessons from other rheumatic conditions indicate that non-standardized recruitment of patients to clinical trials may contribute to the failure of potential novel therapies.

Sarcoidosis remains largely a diagnosis of exclusion. This leaves the fundamental requirement of clinical classification to the discretion of individual investigators. While many recently approved classification criteria for other systemic inflammatory diseases incorporate exclusion elements, these are typically limited to a defined set of alternative diseases or specific clinical or investigational features. For example, the IgG4-related disease criteria exclude fever, eosinophilia > 3000 mm³, cryoglobulinemia, and prominent neutrophilic inflammation, among others, and only four specific diseases [42]. Similarly, the systemic sclerosis criteria exclude 10 scleroderma-like disorders [43], and the SLE criteria rely on attribution rules whereby individual items are not counted if a more likely explanation than SLE exists (e.g., if thrombocytopenia is drug-induced) [44]. In contrast, if the presence of a granuloma were required for sarcoidosis classification, excluding all alternative causes of granulomatous disease would effectively define sarcoidosis almost entirely by exclusion.

This reliance on exclusion highlights a broader limitation of current sarcoidosis frameworks. While they implicitly acknowledge diagnostic uncertainty, they provide little structured guidance on how to actively differentiate sarcoidosis-related manifestations from competing inflammatory, infectious, and neuroimmune conditions. Therefore, diagnostic ambiguity may lead to misattribution of symptoms, delayed recognition of alternative diagnoses, and inappropriate immunosuppression. To address this gap, future classification criteria should transcend passive exclusion and incorporate structured attribution rules and organ-specific “red flags,” thus enabling classification to function as an active aid for differential diagnosis, particularly with clinically complex multisystem disease presentations.

Neurologic involvement illustrates this challenge particularly well. Neurologic symptoms are frequently non-specific and overlap substantially with infectious, inflammatory, vascular, and primary neuroimmune disorders, yet current frameworks offer limited guidance on how such features should be weighted or attributed. This increases the risk of neurosarcoidosis overdiagnosis, especially for patients with systemic sarcoidosis and coincidental neurologic findings. Contextualizing neurologic manifestations within structured attribution rules—emphasizing diagnostic certainty, objective paraclinical evidence, and exclusion of more prevalent alternative diagnoses—could improve specificity, while also preserving sensitivity and reducing inappropriate diagnostic labeling and treatment escalation [45, 46].

Clinical Relevance: Comorbidities and Generalizability

Comorbidities represent a critical yet insufficiently integrated dimension in current sarcoidosis classification frameworks. Cardiovascular, metabolic, infectious, and neurologic comorbid conditions are well-known to influence symptom burden, disease trajectory, therapeutic tolerance, and outcomes [47–51] and should, therefore, not be treated merely as peripheral or confounding factors. In contrast, classification criteria for other systemic inflammatory diseases, such as SLE, already rely on multidimensional, weighted frameworks in which organ involvement, disease activity, and exclusion of comorbidity-driven manifestations are explicitly incorporated to improve specificity and clinical applicability. Importantly, the present work does not aim to deliver a finalized operational framework, but rather to emphasize the need for next-generation classification criteria in sarcoidosis that integrate comorbidities as disease-modifying axes alongside organ involvement, inflammatory activity, and fibrotic burden, to better reflect real-world disease complexity.

While classification criteria may inform clinical reasoning, failure to meet classification thresholds does not exclude a clinical diagnosis of sarcoidosis. In this context, the clinical utility of future classification frameworks will depend on their feasibility and generalizability across care settings, including outside tertiary referral centers. Given the multisystem nature of sarcoidosis, criteria that rely on readily accessible clinical, laboratory, and imaging features and are compatible with an internal medicine–led multidisciplinary assessment, rather than disease-specific specialty silos, are more likely to be broadly applicable. Such an approach would support consistent patient classification for research while preserving diagnostic flexibility in routine clinical practice.

Our initiative to establish classification criteria for sarcoidosis aligns with the growing need for a systematic approach to this complex systemic disease. The American College of Rheumatology (ACR) has developed stringent methodology for development of classification criteria, which have been developed and published [52]. We intend to employ the detailed methodology suggested by the ACR (Table 4).

Table 4.

ACR criteria development checklist (updated March 2018)

1. Will a comprehensive list of possible criteria be considered (content validity)?

2. Will each of the potential criteria be reliable (reproducible), precise in item measurement, easy to measure, and clinically sensible?

3. Are the potential criteria redundant (i.e., highly correlated)? Will this be assessed?

4. Selection of cases (patients considered to have the condition of interest):

a. Will cases be chosen across the spectrum of disease severity?

b. If the criteria are to be used for epidemiologic studies, will both clinical and community cases be included?

5. Selection of controls (patients considered not to have the condition of interest):

a. Will the controls be chosen with a view to the intended purpose of the criteria (i.e., to distinguish individuals with disease from those without disease versus to distinguish individuals with a particular rheumatic disease from individuals with other diseases?

b. Ideally, multiple control groups will be used.

6. Will at least 100 cases and 100 controls be chosen?

7. For each individual criterion, and for combinations of criteria, will the sensitivity and specificity for detecting and ruling out the disease of interest be calculated (construct validity, convergent and divergent validity)? Will these results, together with clinical opinion, be used to reduce the number of criteria for inclusion?

8. Are the criteria to be included those with the greatest content and construct validity? How will this be demonstrated?

9. Will acceptable statistical approaches be used to create the classification criteria from the reduced number of criteria?

10. Will the final classification criteria be validated in different samples of cases and controls? How will those other samples be chosen?

The American College of Rheumatology (ACR) requires project proposals to include a plan for validation of proposed criteria in an independent data set that includes both cases and controls

Conclusions

Sarcoidosis currently lacks validated classification criteria, a gap that undermines trial design, complicates interpretation of existing studies, and slows therapeutic innovation. Building on established ACR/EULAR methodologies used for various diseases such as antiphospholipid syndrome, IgG4-related disease, SLE, and systemic sclerosis, this project will develop and validate research-oriented classification criteria for sarcoidosis. The planned criteria will prioritize specificity, explicitly incorporate mimickers, and combine systemic sarcoidosis classification with weighted organ-involvement domains. By providing standardized entry definitions for sarcoidosis studies, this work aims to reduce cohort heterogeneity, improve the quality and comparability of clinical trials, and ultimately facilitate the development of effective, targeted therapies for patients with this complex multisystem disease.

Supplementary Information

Below is the link to the electronic supplementary material.

Author Contributions

Nadera J. Sweiss: Dr. Sweiss conceptualized and designed the study proposal, led the scoping literature review, developed the search strategy, supervised data extraction, and directed the first draft of the manuscript. She reviewed the manuscript and approved the final version for submission to the journal. Arnaud Dominati: Dr. Dominati helped analyze the concept, and reviewed and approved the final version for submission to the journal. Mary McGowan: The Foundation for Sarcoidosis Research, in part, funded the study. Mary McGowan reviewed and approved the final version for submission to the journal. Tricha Shivas: The Foundation for Sarcoidosis Research, in part, funded the study. Tricha Shivas reviewed and approved the final version for submission to the journal. Alexandra Ladouceur: Dr. Ladouceur reviewed the manuscript and approved the final version for submission to the journal. Israel Rubinstein: Dr. Rubinstein helped analyze the concept, and reviewed and approved the final version for submission to the journal. Murray Baron: Dr. Baron assisted Dr. Sweiss with concept development and the study proposal. He reviewed the manuscript and approved the final version for submission to the journal.

Funding

This study was supported by the University of Illinois College of Medicine, the Bernie Mac Foundation, and the Foundation for Sarcoidosis Research, all in Chicago, IL, USA. The Rapid Service Fee was provided by UIC through these funding sources.

Data Availability

This article is based on previously conducted studies. All source data are contained in cited articles in the References section.

Declarations

Conflict of Interest

Nadera J. Sweiss, Arnaud Dominati, Mary McGowan, Tricha Shivas, Alexandra Ladouceur, Israel Rubinstein, and Murray Baron have nothing to disclose.

Ethical Approval

This article is based on previously conducted studies and does not include any new studies with human participants or animals performed by any of the authors.