Introduction

The disease known widely as Sjögren syndrome affects 1–72 per 10,000 people1 and presents with a diverse spectrum of clinical manifestations2,3. The first cases were reported during the late nineteenth century and were manifested by severe ocular and oral dryness. In 1926, Henri Gougerot4 reported the association of several sicca manifestations in what was probably the first description of the disease. However, there is unanimous agreement in the scientific community that its first comprehensive description as a specific disease was made by the Swedish ophthalmologist Henrik Sjögren in his doctoral thesis published in 1933 (refs. 5,6). In 1936, Stephan Von Grósz7 honoured Sjögren by describing the disease as Sjögren syndrome. The disease received broader recognition after Sjögren’s thesis, which was originally published in German, was translated into English in 1943 (ref. 8). The occurrence of Sjögren syndrome as a ‘sicca complex’ in association with rheumatoid arthritis (RA), systemic sclerosis (SSc) or myositis or as a ‘stand-alone’ disease was highlighted by researchers at the NIH in the early 1960s9. In 1979, Haralampos M. Moutsopoulos et al.10 coined the terms ‘primary’ and ‘secondary’ Sjögren syndrome, considering clinical, serological and genetic characteristics of individuals presenting with sicca-related manifestations, either alone or coexisting with RA; these terms have since been used almost universally to refer to and classify this medical condition. This nomenclature and classification approach has been debated in recent years, as researchers have highlighted concerns about the inaccuracy of referring to a disease as a syndrome11,12 and also the lack of value in distinguishing primary from secondary forms of the disease1,13,14,15,16,17.

During preliminary discussions for the organization of the International Symposium on Sjögren Syndrome (ISSS) in Rome, Italy (7–10 September 2022), two different nomenclature-focused projects emerged: one from the USA (led by A.N.B.) and one from Europe (led by M.R.-C.). These two proposals were merged, and an international Task Force was created with the aim of developing a rational consensus on the nomenclature of Sjögren syndrome (referred to hereafter in this manuscript as ‘Sjögren’) based on the clinical experience of international professionals, current scientific knowledge and the perception and experience of patients.

We report herein the results of a comprehensive process to validate a change in nomenclature for Sjögren, including a bibliometric analysis to determine current terminologies applied to this disease across several languages and countries, a systematic literature review (SLR) to define a rationale for distinguishing primary from secondary forms of the disease, and a Delphi consensus process with respect to disease nomenclature, involving both international professionals and patient groups.

Methods

The convenors (M.R.-C. and A.N.B.) invited international experts with an established record of clinical research in Sjögren to participate in the Delphi consensus process. The Steering Committee (M.R.-C., A.N.B., E.K.A., C. Baldini, H.B., C. Bouillot, S.J.B., T.D., K.M.H., L.L., S.M.L., X.M., S.C.P., V.S., A.G.T., M.d.P.B.-Z., B.A.F.) included eight rheumatologists, two internal medicine specialists, two ophthalmologists, two oral medicine specialists, one paediatrician and two patient representatives. The Working Group (J.-M.A., B.A., M.B., S.C., S.d.V., R.I.F., R. Gerli, R. Giacomelli, J.-E.G., G.H.-M., R.J., A.K., S.K.-K., X.L., S.S.M., H.M.M., W.-F.N., P.O., S.R., M.R., A.S., R.H.S., A.S.-A., V.V., C.V., F.V., M.W.-H.) and the Task Force (see list at the end of the article) included 79 specialists in rheumatology, internal medicine, oral medicine, family medicine, paediatrics, ophthalmology, otolaryngology, genetics and other health professionals from 28 countries (Australia, Brazil, Canada, China, Colombia, Egypt, France, Germany, Greece, Hungary, India, Italy, Japan, Mexico, Netherlands, Norway, Poland, Portugal, Romania, Slovenia, South Korea, Spain, Sweden, Turkey, Uruguay, UK and USA). Members of the Working Group were experts with a strong background in clinical research on Sjögren from centres not represented in the Steering Committee; they contributed to the drafting of the Delphi consensus questions and reviewed the manuscript critically before providing final approval. To ensure patient representation, a worldwide participative process was coordinated by K.M.H. and C. Bouillot through their roles in the US Sjögren’s Foundation and Sjögren Europe, respectively. The opinions of 1,431 patients from 23 countries (Argentina, Australia, Cameroon, Canada, Denmark, Finland, France, Germany, Greece, India, Ireland, Japan, Netherlands, New Zealand, Norway, Poland, Portugal, Romania, South Korea, Spain, Switzerland, UK and USA) were canvassed via e-mail and the cumulative results were then reported to the Task Force.

In the absence of international guidelines or recommendations for a definitive methodological approach to developing a consensus on the nomenclature of a disease, the convenors searched for changes in nomenclature in the field of autoimmune diseases. Among the few precedents found, the reasons for changing the nomenclature of an autoimmune disease were diverse, including the avoidance of eponyms from physicians involved with the Nazi regime (for example, Reiter or Wegener) or the use of terms not acceptable for patients (such as primary biliary ‘cirrhosis’), or to achieve ‘nomenclature symmetry’ with respect to previous nomenclature changes in similar diseases (for example, Churg–Strauss Syndrome)18,19,20,21. The number of experts consulted with respect to the name changes varied widely from four18 to twenty-five19, and the inclusion of patient representatives was only reported in relation to the nomenclature of primary biliary cholangitis21. The methodologies for establishing convergence and agreement were diverse, including a consensus letter (in the case of Reiter’s syndrome)18, a face-to-face meeting20, an e-mail exchange plus a face-to-face meeting19 and online surveys distributed among physicians and patients21. Although many methodological approaches have been developed to achieve consensus on disease classification22, we modelled our approach after those used for nomenclature because we felt that it was important to include the opinion of patients in the process. The convenors allowed themselves a degree of methodological flexibility in gathering international patients’ perspectives, taking into account the clear differences between the design of a study aimed at professionals and one aimed at patients.

The convenors then presented the Steering Committee with a proposed rationale, based on this review of historical methodological precedent, for seeking international consensus on a change in the name for Sjögren and its classification as primary versus secondary based. The Steering Committee in turn agreed to define two main areas of analysis and their corresponding core methodological approaches: a nomenclature area (coordinated by A.N.B., with use of a bibliometric analysis) and a classification area (coordinated by M.R.-C., with use of an SLR).

Bibliometric analysis

Experts in linguistics (A.M., D.Y.) carried out a bibliometric analysis with the aid of professional informationists (J.B., B.T.) to determine the frequency (percentage of usage) and yearly trends in use of the terms ‘Sjögren’, ‘Sjögren’s’, ‘disease’, ‘syndrome’, ‘secondary’ and ‘associated’. Linguistic analyses were performed on three corpora of literature. The first corpus was 34,589 unique medical citations and abstracts (duplicates were removed from the 86,909 total records retrieved) published in journals between 1992 and 2022 and identified from five medical literature databases (MEDLINE via PubMed, Embase, CINAHL, Web of Science and Scopus) using the search terms listed in Supplementary Box 1. This first corpus included non-English-language papers with translated titles and abstracts. The second corpus was scanned books contained in the Google Books n-gram database with the search restricted to 1980–2018. The third corpus was scholarly and popular news, spanning multiple world languages and including encyclopedias, Wikipedia, book titles, journal titles and abstracts and general web matches (using Google Translate to generate possible translations of the various English alternatives and then searching for attestation of those candidate phrases on the web in scientific and medical literature). The analyses utilized linguistically normalized n-gram techniques to enumerate the usage of selected search phrases, both in aggregate and as a trend over time.

Systematic literature review

An SLR is aimed at identifying and evaluating all relevant literature on a topic to draw conclusions about a question. The SLR was carried out to characterize the overlap of Sjögren with each systemic autoimmune disease included in Supplementary Box 2 in terms of frequency and differences in the disease phenotype (epidemiological profile, frequency of ocular and oral dryness, frequency of glandular dysfunction, and histopathological and immunological profiles). The SLR was performed by M.R.-C., M.d.P.B.-Z., S.R. and A.S.-A., using MEDLINE with the MeSH term ‘Sjögren syndrome’ and the search terms included in Supplementary Box 2 with restrictions for date (after 1986), studies (humans) and age of study participants (adults). Other databases, such as EMBASE and Cochrane Library, were also searched. Eligible studies for inclusion were those in which the study population included adults (age > 16 years), used the 2002 American–European Consensus Group or 2016 ACR–EULAR classification criteria for Sjögren23,24 (a retrospective check for the fulfilment of these criteria was allowed for studies published before these years), and used the current classification criteria for the concomitant autoimmune diseases in effect at the time of article publication. We only included individuals who fulfilled the 2002 or 2016 classification criteria as these criteria were developed through international collaboration and were more rigorous than earlier criteria, which did not mandate the presence of positivity for anti-SSA or anti-SSB autoantibodies and/or the presence of a positive minor salivary gland biopsy25. Studies were excluded if the classification criteria used (for Sjögren and/or the autoimmune disease) were not detailed in the methods section or if there was a lack of clinical information about patients with Sjögren overlapping with a second disease (‘overlap Sjögren’). For the analysis, the frequencies of key phenotypic features of individuals with overlap Sjögren were compared with those reported in the Sjögren Big Data Consortium cohort (the largest international cohort of patients with Sjögren classified as ‘primary’)26.

Delphi process

The Delphi technique is a systematic process of surveying a panel of experts to arrive at a group opinion or decision. The Delphi process used here was designed by M.d.P.B.-Z., M.R.-C. and A.N.B. and used the Google Forms platform; the patient groups provided responses via Survey Monkey. All participants were assured as a function of this process that their responses to the surveys would remain confidential. A series of statements were made by M.R.-C. and A.N.B. (Supplementary Box 3), and participants were asked to express their agreement or disagreement with each statement on a 1–5 Likert scale, with 1 representing strong disagreement and 5 representing strong agreement with the statement. For statistical analysis, scores of 1 and 2 were grouped as disagreement, 3 was listed as neutral, and 4 and 5 were grouped as agreement. In the first Delphi round, the task force was asked about their agreement with the use of the following individual terms: the eponym ‘Sjögren’, ‘syndrome’, ‘disease’, ‘primary’, ‘secondary’ and ‘associated’ (Supplementary Fig. 1a). Supplementary Box 3 provides the wording of each statement and its particular pros and cons as discussed by the Steering Committee. A consensus was considered achieved when there was an initial majority in favour (defined as more than two-thirds of participants indicating agreement) or against (more than two-thirds of participants indicating disagreement) each statement. For this first Delphi round, the results from the survey of patients were evaluated quantitatively (arithmetic mean). The results of the first Delphi round were presented and discussed in a plenary session at the 2022 ISSS in Rome to garner the collective opinion of those professionals and patient representatives in attendance. After collecting the arguments for and against each option, the Steering Committee selected five nomenclature options for the second Delphi round. The selection process was conducted entirely online. Steering Committee members were invited to submit their proposed nomenclature options along with a supporting rationale. The convenors compiled all suggestions and subsequently conducted an online survey among the Steering Committee members, asking them to rank their top five options. The five options with the highest rankings were then forwarded to the second Delphi round.

The second Delphi round had two main objectives (Supplementary Box 4). The first objective was to evaluate the degree of support of the scientific community and patient representatives for the usage of each of the five nomenclature options selected by the Steering Committee after collecting feedback from the ISSS meeting in Rome. The number of Delphi rounds for choosing the nomenclature was initially set to two consecutive elimination rounds by seeking a majority opinion (more than two-thirds of participants scoring agreement) (Supplementary Fig. 1b).

The second objective was to assess the majority opinion among task force participants regarding the use of the terms ‘primary’ and ‘associated’ for Sjögren occurring, respectively, alone or in association with another systemic autoimmune disease, by posing three general questions.

Drafting and approval of the manuscript

The manuscript was drafted by M.R.-C., A.N.B. and M.d.P.B.-Z. and was sent sequentially to the Steering Committee and Working Group members for review and approval. The final document is intended to be useful for the international community of health care professionals, doctors and dentists in specialist training, medical students, patients, the pharmaceutical industry and drug regulatory organizations. Industry involvement was not permitted at any stage of the project.

Results

Bibliometric results

As shown in Table 1, ‘Sjögren syndrome’ and its linguistic and spelling variants were the predominant terms for the disease in the literature, accounting for 75% of the identified usage as assessed by an analysis of PubMed citations. In an n-gram analysis of 86,909 medical abstracts of articles (including their titles and keywords) published between 1992 and 2022, the term ‘Sjögren syndrome’ was used 97.8% of the time, with ‘Sjögren disease’ used 1% of the time, ‘autoimmune exocrinopathy’ used 0.4% of the time and ‘Gougerot-Sjögren’ used 0.8% of the time. Among 82,901 medical citations, the possessive form ‘Sjögren’s’ was utilized 72.6% of the time, compared with 26.6% of the time for Sjögren and 0.8% for Sjögrens (including all spelling and linguistic variants for each term). Usage of the possessive form was 82% in the 1993–1997 period but only 70% in the 2018–2022 period, indicating a trend in the published literature away from the use of the possessive Sjögren’s, Sjögren’s and Sjoegren’s (and other spelling variants). There was also a substantial trend over time in the published medical literature away from the usage of ‘secondary Sjögren(’s)’ and its multiple variants with respect to spelling, capitalization, hyphenation and use of the possessive.

Table 1 Synonyms and classification terms for Sjögren and their relative frequencies
Full size table

Systematic literature review

A total of 55 individual studies fulfilled the SLR inclusion criteria27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81 and included a total of 1,931 patients diagnosed with both Sjögren and an associated systemic autoimmune disease (Table 2). Figure 1 summarizes the frequency of Sjögren in people with other systemic autoimmune diseases. The frequency of Sjögren varied widely depending on the associated disease, ranging from 34% in those diagnosed with inflammatory myopathies to 1.7% in individuals with sarcoidosis. The main epidemiological characteristics of patients at the time of Sjögren diagnosis varied substantially depending on the associated disease. The association between Sjögren and some diseases (such as anti-neutrophil cytoplasmic antibody-associated vasculitis, SSc, spondyloarthritis and sarcoidosis) had a substantially reduced female predominance than that reported in Sjögren in the absence of another coexisting disease (‘primary’ form) (93.4% women) (Fig. 2). With respect to age at diagnosis of Sjögren, the coexistence of anti-neutrophil cytoplasmic antibody-associated vasculitis with Sjögren raised this figure from 52 years (reported in the Sjögren Big Data Consortium cohort26) to 62 years, whereas the coexistence of systemic lupus erythematosus (SLE) lowered the age at diagnosis of Sjögren to below 45 years (Supplementary Fig. 2). There were notable differences in the frequency of glandular involvement (Supplementary Figs. 3 and 4) and the immunological profile (Supplementary Fig. 5) in comparison with what was reported in the Sjögren Big Data Consortium cohort26.

Table 2 Key features of patients with Sjögren and an associated systemic autoimmune disease
Full size table
Fig. 1: Frequency of Sjögren disease among individuals with another concomitant systemic autoimmune disease.
figure 1

The figure summarizes the frequency of Sjögren disease (SjD) in the studies included in a systematic literature review27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81; the 55 studies that met the inclusion criteria involved a total of 1,931 patients diagnosed with both Sjögren and an associated systemic autoimmune disease (see Table 2). The frequency of SjD varied widely depending on the associated disease. RA, rheumatoid arthritis; SSc, systemic sclerosis; SLE, systemic lupus erythematosus; SpA, spondyloarthritis.

Full size image
Fig. 2: Frequency of women among patients with Sjögren disease and other systemic autoimmune diseases.
figure 2

The radar chart depicts the percentage of patients with Sjögren disease and another concomitant systemic autoimmune disease who are women. The values are derived from those studies included in the systematic review that provided sex-specific patient information27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81; the 55 studies that met the inclusion criteria involved a total of 1,931 patients diagnosed with both Sjögren and an associated systemic autoimmune disease (see Table 2). In the figure, the label ‘primary Sjögren’ corresponds to patients classified as having ‘primary’ Sjögren (that is, Sjögren in the absence of another coexisting disease) in the Sjögren Big Data Consortium cohort26. AAV, anti-neutrophil cytoplasmic antibody-associated vasculitis; IgG4-RD, IgG4-related disease; RA, rheumatoid arthritis; SSc, systemic sclerosis; SLE, systemic lupus erythematosus; SpA, spondyloarthritis.

Full size image

In summary, the results generated from the SLR demonstrate the enormous influence of the association with another systemic autoimmune disease, with substantial alteration of both the epidemiological profile and the phenotype of Sjögren compared with the characteristics observed in the absence of another coexisting disease.

Delphi process on terms to name the condition

In the first Delphi round, which assessed the extent of agreement on the use of individual terms, 88% of the Task Force participants agreed with the statement about maintaining the use of the eponym ‘Sjögren’, 31% agreed with the statement about maintaining the term ‘syndrome’ and 53% agreed with the statements about changing the term ‘syndrome’ to ‘disease’. With respect to patients’ opinions, on a numeric scale on which 5 was the maximum level of agreement, mean scores were 4.35 for maintaining the use of the eponym ‘Sjögren’, 1.87 for using the term ‘syndrome’, and 4.23 for using the term ‘disease’ (Supplementary Table 1).

After discussion online among the Steering Committee members, we opted to forego any effort to reach consensus on the use of the possessive or non-possessive form of Sjögren in English publications, or of spelling variants (o, oe, ö), and left this up to individual or journal preference.

In the second Delphi round, which assessed the extent of agreement with the use of the five proposed nomenclature terms, 63% of the task force participants supported the use of ‘Sjögren disease’ (with the abbreviation SjD), 28% supported the use of a novel acronym ‘S.J.Ö.G.R.E.N.’ (salivary gland swelling, joint pain/swelling, ocular/oral dryness, general symptoms, renal/respiratory manifestations, exocrinopathy, non-Hodgkin lymphoma), 22% supported the use of ‘Sjögren syndrome’, 12% supported the use of ‘Sjögren autoimmune epithelitis’ and 5% the use of ‘Sjögren autoimmune exocrinopathy’. Among the patients, 77% supported the use of ‘Sjögren disease’ followed by 25% supporting the acronym S.J.Ö.G.R.E.N. (Supplementary Table 2). After combining the results of the two arms, 70% of participants (professionals plus patients) supported the use of the term Sjögren disease, avoiding the need for a subsequent Delphi round because a majority opinion had been obtained (Fig. 3).

Fig. 3: Agreement on the use of proposed nomenclature terms for Sjögren.
figure 3

The figure summarizes the results of the second Delphi round, in which Task Force participants and patient representatives were surveyed regarding their support for the usage of each of five nomenclature options. Participants were asked to express their agreement or disagreement with the use of each term on a 1–5 Likert scale, with 1 representing strong disagreement and 5 representing strong agreement; results of the survey are provided in Supplementary Table 2. Overall, a majority of participants supported the use of the term ‘Sjögren disease’.

Full size image

Delphi process on classification nomenclature

In the first Delphi round, which assessed the extent of agreement by clinicians on the individual terms used in the classification of Sjögren, there was a similar level of support for either maintaining the term ‘primary’ (44%) or not maintaining it (41%), whereas a majority (73%) agreed on not maintaining the term ‘secondary’. If the term ‘secondary’ was not maintained, 75% agreed on using the alternative term ‘associated’. Patients’ opinions were quantified on a numeric scale on which 5 was the maximum level of agreement; the mean score was 2.51 for maintaining the use of the term ‘primary’, 2.15 for maintaining the term ‘secondary’ and 3.08 for changing the term ‘secondary’ to ‘associated’ (Supplementary Table 1).

In the second Delphi round, those surveyed were asked if there were objective reasons to differentiate between primary and associated Sjögren for disease nomenclature, specifically given that this distinction is not applied to other systemic autoimmune diseases (Supplementary Table 2), which do not always occur in isolation. Most professionals (76%) and patients (79%) agreed that there were no reasons to apply this differentiation specifically to Sjögren when it is not applied to the autoimmune diseases commonly associated with Sjögren. However, 54% of professionals and 44% of patients replied that the distinction between primary and associated Sjögren still had potential importance, with 30% of professionals stating that this importance is contextual (as indicated by the response “It depends”). When professionals were asked in which field it is necessary to differentiate between primary and associated Sjögren, 50% indicated mainly in the scientific field (for the purpose of defining a study population) and 25% indicated in all fields, including in clinical practice for individual diagnosis. Overall, 82% of experts and 55% of patients agreed that it was important to differentiate primary Sjögren from associated Sjögren for scientific studies, clinical practice or both. These results were discussed online among the Steering Committee members (including patient representatives), and it was agreed with the patient representatives to formulate two recommendations that reflected an agreement position (recommendations 4 and 5 in Box 1). These recommendations were sent electronically to the entire task force and were approved by all Task Force members after making exclusively minor semantic (not substantive) changes. The survey did not specifically address the need of clinicians to differentiate between primary and associated disease for clinical decision making, such as the attribution of specific systemic features to Sjögren or an overlap systemic disease and the choice of specific treatments.

Consensus was not reached with respect to the importance of the differentiation between primary and associated Sjögren in the field of research. As evidenced by the SLR, the Sjögren phenotype is substantially influenced by the association with other systemic autoimmune diseases (Figs. 1 and 2; Supplementary Fig. 2–5). Thus, it is recommended that researchers detail in future scientific publications on Sjögren whether patients with other autoimmune diseases (either systemic or organ-specific) were excluded from studies (detailing the excluded diseases if so) or, if they were not excluded, the percentage of patients who had other autoimmune disease(s) (whenever possible). Furthermore, it might be necessary to conduct a separate sensitivity study and subgroup analysis, especially in experimental or interventional studies. The scientific rationale for this recommendation is to maintain the comparability of results from new studies with those carried out in the past (practically all of which were carried out in people classified as having ‘primary’ Sjögren).

Discussion

The recommendations of the 2023 International Rome consensus for the nomenclature of Sjögren are summarized in Box 1. We have obtained a clear international consensus that the medical condition known as Sjögren syndrome should be re-named and hereafter referred to as Sjögren (or Sjögren’s) disease or with the abbreviation SjD. This change in nomenclature has been reached and endorsed via a Delphi process with the participation of 79 professionals and 1,431 patients from 34 countries. There are two key arguments in favour of this change in nomenclature. First, it is important to recognize that SjD is not a syndrome, namely an aggregate of symptoms and signs that are associated with a morbid process independent of pathogenesis82,83. Instead, SjD is widely accepted as a distinct autoimmune disease, with characteristic autoantibodies, glandular histopathology and a specific pattern of systemic involvement. Some members of the Task Force thought that changing the term from ‘syndrome’ to ‘disease’ could have positive outcomes regarding research and funding attention. The second argument emerged principally from patients’ beliefs that the term ‘syndrome’ can be counterproductive, indicating that SjD is a loose collection of ill-defined or ‘nuisance’ symptoms (overwhelmingly sicca) rather than the actuality, which is that it is a disease with serious morbidity and, for some, increased mortality84.

Regarding the use of the eponym Sjögren, most patients and professionals supported maintaining its use. Although some experts argued strongly that the eponym should be abandoned in favour of a histopathological descriptor (for example, autoimmune epithelitis), patient representatives argued that it would be very difficult for their advocacy organizations to promote awareness and education and obtain increased funding for research with such a technical disease name. In addition, ‘autoimmune epithelitis’ is hard to remember and does not convey to the public or to patients what the disease is any more than ‘Sjögren’ does.

Finally, the choice of a disease abbreviation (SjD) was also important, not only for conformity but also to ensure that it does not have negative connotations. Until now, the abbreviation SS (commonly used for Sjögren syndrome) applies to 337 terms85, including some with potentially offensive connotations. Similarly, the abbreviation SD applies to over 209 terms, including the medical terms sudden death, senile dementia and sexual dysfunction. By contrast, SjD is more specific and has only one other medical connotation (sacroiliac joint dysfunction).

A clear consensus was also reached on abandoning the term ‘secondary’ in favour of ‘associated’ when classifying patients with SjD co-occurring with another systemic autoimmune disease. An issue that remained without clear resolution was whether there was importance, value or utility in differentiating primary and associated forms of SjD. Nearly 80% of professionals and patients agreed that there are no reasons to label a given patient as having either primary Sjögren or associated Sjögren, particularly as these labels are not applied to patients with SLE, SSc or RA that is ‘stand-alone’ or overlapping with a second disease. Sjögren should be considered a disease in its own right, rather than a secondary manifestation of another disease. The patient representatives made it clear that the use of the terms ‘secondary’ and ‘associated’ gives the impression that their Sjögren disease is less important than other concomitant conditions. However, a majority of professionals and patients appreciated the necessity of distinguishing primary versus associated Sjögren in research studies or clinical practice, or both.

The coexistence in a single patient of more than one autoimmune disease, both systemic and organ-specific, is a well-recognized clinical occurrence with a reported frequency ranging between 8% and 53%86,87,88,89,90,91. In fact, SjD is the systemic autoimmune disease that most commonly overlaps with another92,93. The terms ‘primary’ and ‘secondary’ are misleading when applied to SjD, as they suggest that primary disease is idiopathic whereas secondary disease is derived from a definable related condition. Instead, the co-existence of two or more autoimmune diseases, termed polyautoimmunity, can simply reflect shared genetic and environmental predisposing factors and pathogenetic pathways94,95. In addition, some patients might present with an illness best characterized as SLE or RA but that evolves over time into one best characterized as SjD. The absence of an objective criterion to help differentiate between primary and secondary SjD has resulted in complex issues of classification. Thus, the term ‘secondary’ has been overwhelmingly used when SjD is associated with RA, SSc or SLE, whereas it has been rarely used when it is associated with other systemic autoimmune diseases (including antiphospholipid antibody syndrome, sarcoidosis, myopathies, vasculitis and IgG4-related disease). Why this differentiation is applied to SjD and not to the other systemic autoimmune diseases is unclear but might relate to the initial perception by rheumatologists that SjD was a variant form of RA96. The fact that SjD could occur alone, in the absence of a second autoimmune disease, was first highlighted in landmark studies at the NIH in the early 1960s9. The distinction between primary and secondary forms of the disease was advanced by Moutsopoulos et al.10 and predicated on the concept that each form had unique clinical, serological and genetic characteristics that could influence disease prognosis and appropriate therapeutic strategies. However, in 2019 Mavragani and Moutsopoulos wrote that splitting SjD into ‘primary’ and ‘secondary’ forms fails to fully reveal the wide clinical spectrum of the disease14. The distinction between primary and associated SjD reflects only the frequently reported clinical situation of the coexistence or overlap of SjD in patients with other autoimmune diseases.

An issue that was not specifically addressed in the Delphi process was the value of defining whether specific systemic manifestations of SjD relate to an associated condition, such as a myositis syndrome for patients with interstitial lung disease or neuromyelitis optica or multiple sclerosis for patients with central nervous system demyelination. In these examples, diagnosis of the associated condition might influence a clinician’s choice of therapy. Clinicians often judge whether a patient has primary or associated SjD on the basis of the predominant phenotype, an assessment that can change during the natural history of the patient’s disease. Recognizing a clinical overlap could be valuable as it could influence how we can manage the patient (for example, when interpreting new symptoms, ordering organ-specific tests during follow-up or even choosing a different therapeutic approach). The overlap of two systemic autoimmune diseases can also affect overall disease burden and severity91. This clinical distinction is increasingly murky and awaits advances in new classification tools and molecular markers for clarification.

The distinction between ‘primary’ and ‘secondary’ forms of SjD has had substantial consequences for research, as most published scientific studies in SjD focus on the primary form, excluding patients with an associated systemic autoimmune disease. This differentiation has been maintained until now in both the current classification criteria and activity rating scales. The developers of the current ACR–EULAR classification criteria stated that they “focus on primary rather than secondary [Sjögren syndrome]. Patients with the latter would typically not be eligible for experimental treatments for [Sjögren syndrome]”24. Consistent with this statement, therapeutic studies on SjD, especially randomized controlled trials, systematically exclude patients with other associated systemic autoimmune diseases. Why this occurs only in trials carried out in people with SjD and not in those affected by other systemic autoimmune diseases is difficult to explain from a methodological point of view. Post hoc analyses have even investigated why therapeutic responses can be different in patients with and without associated SjD97,98.

Our SLR demonstrated that the coexistence of another systemic autoimmune disease heavily influences the phenotype of SjD. Thus, the inclusion of patients with both primary and associated SjD in a clinical study or therapeutic trial can alter the homogeneity of the study population and affect patterns of results. In cases in which study design allows for the inclusion of patients with SjD who have associated diseases, it would seem reasonable to conduct a specific sensitivity analysis excluding these patients.

Conclusions

The 2023 Rome International Consensus on Sjögren Nomenclature strongly endorses the transition from the term ‘Sjögren syndrome’ to ‘Sjögren disease’ (SjD) to better reflect its definition as a distinct systemic autoimmune disease. The term ‘associated’ is recommended instead of ‘secondary’ to indicate Sjögren when it is occurring with another systemic autoimmune disease for which classification criteria are fulfilled. It may be unnecessary to label patients with primary versus associated Sjögren in clinical practice, as this distinction is not routinely applied to other systemic autoimmune diseases. However, this recommendation does not negate the potential value of recognizing the overlap of SjD with other systemic diseases in individual patients and its impact on the attribution of systemic manifestations and treatment choices. Similarly, this distinction is useful for research purposes to ensure a uniform study population. This consensus marks a significant stride towards a more streamlined and precise understanding of SjD within both clinical and research settings.