Undifferentiated systemic rheumatic (connective tissue) diseases and overlap syndromes

INTRODUCTION — As much as 25 percent of rheumatic disease patients with systemic symptoms cannot be definitively diagnosed. Furthermore, the majority of these patients will remain undiagnosed during 5 to 10 years of follow-up [1-3].

Such patients were historically described as having "collagen" or "connective tissue" diseases, since they shared similar clinical and pathologic features of widespread inflammation [3-6]. Subsequently, these patients have been considered to have "diffuse" or "undifferentiated" (or sometimes "early undifferentiated") "connective tissue diseases" (UCTD) [3,7-12]. Other patients with rheumatic diseases have systemic features that overlap two or more specific, recognized entities and also cannot be definitively diagnosed [2,3,13-15].

Increasingly, the term "systemic rheumatic" disease has replaced "collagen," "collagen-vascular," or "connective tissue" disease. This change reflects the recognition that these disorders probably represent perturbations of the immune system with resultant inflammatory tissue injury rather than primary disorders of collagen, vasculature, or connective tissue; the inciting events are not yet clear. We will retain both "undifferentiated systemic rheumatic" and "undifferentiated connective tissue" diseases in our discussion.

This topic review provides an overview of undifferentiated systemic rheumatic diseases and overlap disorders, together with an overall approach to these patients.

CLASSIFICATION ISSUES — The American College of Rheumatology (ACR) and the European Alliance of Associations for Rheumatology (EULAR; formerly known as European League Against Rheumatism) have developed classification criteria to assist in the categorization and diagnosis of patients with rheumatic diseases. These classification criteria do not substitute for identification of unique, defining etiologic, clinical, or immunopathologic features for systemic rheumatic diseases, which remain unknown for most of these disorders.

The disorders generally considered to be undifferentiated systemic rheumatic diseases (USRD) and/or overlap syndromes are presented in the table (table 1). Undifferentiated rheumatic diseases generally comprise one or more of the following clinical scenarios:

●Early Raynaud phenomenon alone. (See 'Raynaud phenomenon' below.)

●Early inflammatory (poly)arthritis that does not fulfill ACR criteria for the diagnosis of rheumatoid arthritis (RA). (See 'Polyarthritis' below.)

●Nonspecific rash resembling the cutaneous findings found in defined rheumatic diseases, often associated with an interface dermatitis (ie, inflammation at the dermal-epidermal junction). (See 'Cutaneous lesions' below.)

●Patients who, despite manifesting multiple nonspecific clinical or serologic abnormalities, do not meet established classification or diagnostic criteria for a specific rheumatic disease. These might include certain other manifestations of inflammatory myopathy, systemic sclerosis, systemic lupus erythematosus (SLE), Sjögren's disease, inflammatory arthritis, vasculitis, serositis, or interstitial lung disease which fall short of meeting acceptable diagnostic criteria for a single entity. (See 'Early undifferentiated systemic rheumatic disease' below.)

●Interstitial lung disease and/or nonspecific interstitial pneumonia are generally recognized as either possible initial manifestations of or associated with undifferentiated systematic rheumatic disease. (See 'Nonspecific interstitial pneumonia' below.)

Diagnostic caution — We urge caution in assigning a diagnosis, or even offering several diagnoses (eg, RA and SLE), for patients with symptoms of rheumatic diseases who do not meet currently accepted criteria [3,16-19]. Assigning diagnostic labels to patients implies certainty or confidence in pathogenesis, prognosis, therapeutic implications, and outcomes. Diagnosis should not be based only on certain laboratory or serologic abnormalities or on isolated clinical features. Rendering a diagnosis in such circumstances may inappropriately circumscribe the thinking of the health care providers, and may lead to inappropriate management and harm. There may also be legal consequences from such mislabeling. In circumstances of diagnostic uncertainty, we refer to such patients as having "undifferentiated systemic rheumatic disease" [3]. (See 'General approach to the patient' below.)

Raynaud phenomenon — The Raynaud phenomenon (RP) may be an early sign of systemic rheumatic disease in as much as 10 percent of individuals. As an example, in one study of patients initially presenting with RP, 36 percent evolved into undifferentiated connective tissue diseases (UCTD). Thumb involvement with RP may be more likely with secondary than primary RP [20].

Patients with RP should be evaluated with a careful history and physical examination, including nailfold capillaroscopy (which may be done in the office with a handheld lighted microscope (magnifier) or with an ophthalmoscope and immersion oil), testing for antinuclear antibodies (ANA), and, in selected patients, testing for the presence of cryoglobulins or other cryoprecipitable proteins. As RP is a common presenting feature of antisynthetase syndrome, consideration of a possible inflammatory myopathy may be warranted. (See "Clinical manifestations and diagnosis of Raynaud phenomenon" and "Overview of and approach to the idiopathic inflammatory myopathies", section on 'Antisynthetase syndrome'.)

Nailfold capillaroscopy — The value of abnormal nailfold capillaroscopy (NC) in predicting the development of a defined rheumatic disease was emphasized by a study of 1024 patients with RP, among whom 308 had neither serological findings nor clinical signs of rheumatic disease but did have NC pathological features suspicious for rheumatic disease [21]. Follow-up data were available for 133 patients. After a mean follow-up of 6.5 years, 109 patients had developed a rheumatic disease (mostly diffuse and limited cutaneous systemic sclerosis, sometimes with findings consistent with the CREST syndrome [calcinosis, RP, esophageal dysmotility, sclerodactyly, and telangiectasia syndrome] or mixed connective tissue disease [MCTD]), and 24 patients did not show any clinical signs or serological markers for a rheumatic disease after a mean follow-up of 8.5 years. The presence of giant capillaries, avascular fields, and irregular architecture of nailfold capillaries was highly predictive for the development of a rheumatic disease in patients with RP.

In addition to positive findings on NC, other features predictive of underlying systemic rheumatic disease included older age of onset, severe digital ischemia, asymmetrical attacks, and the presence of ANA [22]. The presence of ANA alone, however, should not lead to a premature diagnosis of underlying systemic rheumatic disease. In a prospective study of 1039 patients with RP, there were 129 who had a positive ANA among 819 subjects who had no systemic rheumatic disease at initial presentation [23]. A defined rheumatic disease developed in only 22 percent of this subset after a mean of three years of follow-up.

In contrast to the lack of predictive value of ANA, the presence of anticentromere antibodies was suggestive of the eventual development of CREST syndrome. In one series of 85 patients with RP, the presence of anticentromere antibodies had a sensitivity of 60 percent and specificity of 98 percent for the development of CREST among those presenting with primary RP or with possible connective tissue disease [24].

Polyarthritis — It is important to confidently diagnose RA early in its course in order to identify those patients who might benefit from early aggressive therapy. This follows from an appreciation that irreversible joint damage can occur within the first six months of onset of RA and that a narrow window of opportunity exists in which to optimally introduce effective antirheumatic therapy. (See "Diagnosis and differential diagnosis of rheumatoid arthritis" and "General principles and overview of management of rheumatoid arthritis in adults" and "Undifferentiated inflammatory arthritis in adults".)

For patients who meet ACR criteria for classic RA when first evaluated, approximately 90 percent carry the same diagnosis five years later [12,25].

At least as common as RA is a heterogeneous group of patients with the following findings:

●Symptoms of inflammatory joint disease such as diffuse or migratory arthralgias, symmetrical or asymmetrical polyarthritis or oligoarthritis, migratory polyarthritis, a convincing history of joint swelling, or palindromic symptoms [26].

●A history which frequently includes prolonged morning stiffness

●An elevated erythrocyte sedimentation rate and positive ANA or rheumatoid factor tests

●A dramatic response to antiinflammatory therapy [27,28]

These findings are suggestive of RA but do not fulfill ACR criteria for this diagnosis. Approximately 10 percent of such patients will ultimately meet diagnostic criteria for another systemic rheumatic disease (lupus, systemic sclerosis, polymyositis, or a spondyloarthropathy), and 5 to 10 percent will acquire other features of undifferentiated systemic rheumatic disease but will still not be diagnosable as having a specific rheumatic disease.

In two large prospective studies, approximately 20 percent of such patients evolved into definite RA [12,25]. A consistent pattern of predictive variables could not be identified in the multivariate analyses other than higher small joint counts and higher erythrocyte sedimentation rate at year one in one study [12] and rheumatoid factor positivity in the other [25].

Another study of 1030 patients with early RA including 435 who were seronegative found that there was significant reclassification and diagnostic heterogeneity in these patients during the 10-year follow-up period [29]. Three percent of the seronegative cases were reclassified as seropositive or erosive RA, 16 percent as polymyalgia rheumatica, 11 percent as psoriatic arthritis, 10 percent as osteoarthritis, 9 percent as spondyloarthritis, 3 percent as reactive arthritis, 2 percent as gout, 4 percent as pseudogout, 1 percent as paraneoplastic arthritis, 1 percent as juvenile arthritis, and less than 1 percent each as hemochromatosis, ankylosing spondylitis, giant cell arteritis, and other miscellaneous diagnoses.

The prognosis is generally considered different between those with undifferentiated arthritis (UA) and those with RA. The following are illustrative:

●In one report of 1141 such patients (638 with UA and 503 with RA), complete resolution of symptoms occurred in 54 percent of those with undifferentiated polyarthritis versus only 7.5 percent at seven years in those with RA at presentation [25].

●In contrast, a Dutch study of the one-year outcome of patients with UA found that 42 percent of patients had progressive disease, most often identified among those with older age, higher disease activity, and arthritis of the hands at baseline. With the availability of potent medications to arrest the rheumatoid process, the authors concluded that treatment should be directed towards the severity of the presentation rather than the diagnostic categorization [30].

●Similar outcomes were seen for patients presenting with either RA or UA evolving into RA [31].

Early intervention with a disease-modifying antirheumatic drug (DMARD) may alter the natural progression from UA to RA [32]. This was illustrated in a study that randomly assigned 110 patients with UA to either methotrexate (MTX) or placebo [33]. Dose was adjusted based upon disease activity score and was then tapered and discontinued after 12 months. Progression from UA to RA occurred in a significantly smaller proportion of those receiving MTX than placebo (40 versus 53 percent, respectively). Fewer MTX treated patients showed radiographic progression over 18 months. Certain clinical features (sex, age, morning stiffness, localization of symptoms, tender and swollen joint counts, C-reactive protein, rheumatoid factor positivity, and presence of anticyclic citrullinated protein/peptide [CCP] antibodies) predicted risk of developing RA and, therefore, could allow individualization of therapeutic decisions for patients with early, undifferentiated arthritis [34].

Among patients with early undifferentiated arthritis, serologic testing for anti-CCP has been demonstrated to be valuable in predicting progression to RA. This was illustrated in a prospective study of 346 patients with arthritis of less than two years' duration who did not meet classification criteria for any specific type of arthritis [35]. In this group of patients, 69 were anti-CCP antibody positive, and 64 of them (93 percent) met classification criteria for RA in the subsequent three years. In contrast, among the 249 patients who were initially anti-CCP negative, 63 (25 percent) developed sufficient criteria to establish the presence of RA during the same follow-up period. (See "Biologic markers in the assessment of rheumatoid arthritis", section on 'Anti-citrullinated peptide antibodies'.)

Unconventional serologic findings (eg, IgA rheumatoid factor, pyridinoline) have also been noted to help predict the later development of erosive disease [36].

The use of ultrasound or magnetic resonance imaging (MRI), which offer the ability to identify synovitis, bone edema, and erosions prior to the development of classical x-ray changes, may help separate those patients likely to develop RA from patients with less aggressive inflammatory joint disease. An algorithm including MRI-proven early joint damage in conjunction with autoantibodies has been promoted to predict progression from UA to RA [37]. A study using MRI found more destructive change (synovial thickening, marrow edema, and erosions) in patients with RA and more enthesopathy in patients with spondyloarthropathy than in patients with undifferentiated arthritis [38]. In patients with undifferentiated spondyloarthropathies, positive testing for HLA-B27 and the presence of buttock pain predicted evolution to a definable disease [39]. While HLA-B27 was slightly more common in patients with UA than controls (10.1 versus 7.2 percent), it was suggested that because of the low yield of HLA-B27 testing in this context, it be reserved for patients with other features of spondyloarthropathy [40].

With the revisions incorporated in the 2010 ACR/EULAR RA classification criteria, including the presence of anti-CCP antibodies, it was anticipated that the diagnosis of RA would be entertained in individuals with much earlier presentations [41,42]. This has been borne out by studies documenting that the UA is less prevalent and milder in its presentation and outcome since the 2010 criteria were instituted, suggesting that patients with unfavorable characteristics are more frequently classified as having RA [43]. (See "Diagnosis and differential diagnosis of rheumatoid arthritis".)

Cutaneous lesions — The cutaneous manifestations of UCTD are not well defined. In one study of 526 individuals with UCTD and antithyroid antibodies, 47 patients had a dermatitis which consisted of erythematous macules, patches, or papules with delicate scale [44]. The skin lesions were found most frequently on the lower extremities, but were also observed on the upper extremities and back. Skin biopsies in 15 of 47 of these individuals showed mild lymphocytic dermatitis, superficial perivascular infiltrates of mononuclear inflammatory cells that spared eccrine glands, patchy dermal mucin, and occasional lymphocytic vasculopathy. Immunofluorescent staining was not performed. Of interest to the authors of the study, this form of dermatitis was more common in patients with antithyroid antibodies (9 percent) than in those without such autoantibodies (2 percent).

Nonspecific interstitial pneumonia — Interstitial lung disease (ILD, also known as idiopathic interstitial pneumonia [IIP]) may precede or occur at or after onset of various systemic and rheumatic disorders. A histologic pattern referred to as nonspecific interstitial pneumonia pattern (NSIP) is commonly found in lung biopsy specimens of patients with many rheumatic and systemic diseases, including SLE, Sjögren's disease, systemic sclerosis, and dermatomyositis. Interestingly, the histologic pattern of usual interstitial pneumonia (UIP), associated with a poorer prognosis, is the more common presentation of ILD in association with RA [45]. (See "Treatment and prognosis of nonspecific interstitial pneumonia".)

The NSIP pattern of lung injury has also been noted to be a common pattern in patients who are classified as having UCTD. This was illustrated in a case-control study of 28 patients with one or more clinical features of USRD and ILD and a control group of 47 patients with ILD in the absence of USRD [46]. Subgroups of USRD and control patients without clinical features of USRD underwent lung biopsy (18 and 22 patients, respectively). The histologic pattern of NSIP was most common in those with USRD (83 percent), with other patterns including usual interstitial pneumonia (UIP), organizing pneumonia, and nonclassifiable fibrosis less often noted (6 percent each). In contrast, in patients with IIP and no clinical features of USRD, UIP was present in 86 percent, and NSIP was present in only 9 percent. In a series of 68 patients who had presented with IIP and who were followed prospectively for 11 years, 13 (19 percent) eventually developed classifiable rheumatic disease [47].

It was uncertain whether patients with USRD and ILD were more or less likely to develop a classifiable systemic or rheumatic disease. In the study cited above [46], patients with lung involvement had pulmonary symptoms such as cough or dyspnea for a mean of 3.5 years before the time that features of USRD were assessed.

The concept of "lung-dominant CTD" has been proposed to characterize those patients with NSIP who had certain histopathologic features (diffuse perivascular collagen, aggressive pleuritis, lymphoid aggregates with germinal center formation, and/or prominent plasma cell infiltrates) or positive serologies (antisynthetase, Scl-70, SSA, or CCP antibodies) but who lacked sufficient extra-pulmonary clinical features to be classified as having a specific rheumatic disease. This may be an important distinction, since the prognosis and response to therapy for rheumatic disease-associated NSIP are more favorable than for IIP that is not associated with rheumatic disease [48,49]. Even in patients with UIP, the search for signs of an underlying rheumatic disease may be warranted because of the significantly better survival of UIP associated with rheumatic disease than IIP that is not associated with rheumatic disease [45]. The presence of myositis-specific antibodies such as those associated with antisynthetase syndrome or melanoma differentiation-associated gene 5 (MDA-5) may play a role in further characterizing lung-dominant CTD patients, especially those who do not have a classic myositis phenotype.(See "Overview of and approach to the idiopathic inflammatory myopathies", section on 'Myositis-specific autoantibodies'.)

The European Respiratory Society/American Thoracic Society has proposed the term "interstitial pneumonia with autoimmune features (IPAF)" for patients with ILD and features of autoimmunity [50]. Others have attempted to distinguish IPAF from UCTD, finding that UCTD patients were more likely to be female and younger than patients with IPAF or ILD [51]. The value of efforts to dissect unclassifiable illnesses without a clear understanding of etiologies, pathogenesis, outcomes, and/or therapeutic responses is uncertain.

Early undifferentiated systemic rheumatic disease — Patients who do not present with a clearly defined autoimmune/rheumatologic syndrome or who have overlapping features of two or more diseases have been estimated to comprise 15 to 25 percent of tertiary rheumatology referrals [1,2]. One of the difficulties of making a specific diagnosis of a discrete systemic rheumatic disease is that several clinical features may be shared by RA, Sjögren's disease, SLE, systemic sclerosis, and inflammatory myopathy. These include RP, polyarthritis, ILD, pleuritis, or, less often, pericarditis and small vessel vasculitis.

Another part of the diagnostic problem is that these various diseases may share serological abnormalities. This is particularly true for ANA, rheumatoid factor, anti-Ro/SSA, and anti-La/SSB. Some antibodies, however, have greater diagnostic value. These include anticentromere and anti-topoisomerase-I (Scl-70) antibodies in limited cutaneous systemic sclerosis (CREST syndrome) and diffuse cutaneous systemic sclerosis, respectively; anti-CCP antibodies in RA; and anti-Sm, anti-double-stranded DNA (dsDNA), and antiphospholipid antibodies in lupus [52]. There remains a need for identifying meaningful preclinical biomarkers in patients with systemic autoimmune rheumatic disease [53,54]. (See "Measurement and clinical significance of antinuclear antibodies".)

It was found that ultrasonography of the major salivary glands helped distinguish patients with primary Sjögren's disease from those with undifferentiated systemic rheumatic disease. Patients with Sjögren's disease had a salivary gland ultrasonography score of 2.2 versus 0.2, on average, for undifferentiated systemic rheumatic disease, with a sensitivity of 65 percent, specificity of 96 percent, positive predictive value of 95 percent, and negative predictive value of 73 percent for the diagnosis of Sjögren's disease [55]. (See "Diagnosis and classification of Sjögren’s disease".)

Patients presenting with incompletely expressed rheumatic disease and nonspecific autoantibodies are designated as having undifferentiated systemic rheumatic (connective tissue) disease. A premature diagnosis of one of the discrete rheumatic diseases does not serve patients well, since the undifferentiated nature of the syndrome usually persists. In general, among those destined to develop a defined systemic rheumatic disease, evolution occurs early in the course of the presentation, within two to five years in the majority of cases [56].

Patients with "early," "incomplete," or "preclinical" autoimmune systemic rheumatic disease had increased risk of fetal growth restriction preeclampsia and small-for-gestational-age infants when compared with controls who were without any rheumatologic symptoms or features [57]. Another series of 100 pregnancies in patients with UCTD reported 11 first-trimester miscarriages, 89 live births, and 13 flares during pregnancy or puerperium, 3 of which were major and led to development of SLE with renal disease [58]. There were 26 obstetrical complications among the 89 successful pregnancies (29 percent), including 1 of preeclampsia. A significant link between anti-dsDNA at baseline and disease activity at the beginning of pregnancy was noted. The impact of pregnancy on disease in these patients was thought to be less than for patients with other rheumatic diseases.

Evolution from undifferentiated systemic rheumatic (connective tissue) disease to a discrete rheumatic disease is somewhat more common than resolution of the syndrome. In general, a definitive rheumatic disease is diagnosed in 10 to 35 percent of patients 3 to 10 years from disease onset [12,56,59-64]. Illustrative results of prospective studies include the following:

●In a North American series of 143 patients followed prospectively for up to 10 years, a definite rheumatic disease was observed in 29 percent of individuals (SLE in 13 percent, systemic sclerosis in 15 percent, RA in 3 percent, and MCTD in 3 percent), while complete remission occurred in 6 percent [59,60].

●In a series of 665 Hungarian patients followed for five years, evolution to a well-defined autoimmune syndrome occurred in 35 percent of patients, with 28 developing SLE, 26 MCTD, 19 systemic sclerosis, 45 Sjögren's disease, 3 poly-/dermatomyositis, 87 RA, and 22 "systemic vasculitis," while 65 percent remained in the UCTD category [62]. Among all patients in this study, complete remission occurred in only 12 percent.

●A multicenter study evaluated a cohort of 213 patients with symptoms of early undifferentiated systemic rheumatic disease for at least one year who were then followed for five years [12]. The rate of progression to RA or SLE was 20 and 13 percent, respectively.

●Another study reported outcomes from 98 patients with UCTD who were followed for a mean of 11 years [65]. Investigators found that 62 percent of patients continued to have UCTD, 24 percent entered remission, and 14 percent developed definite rheumatic diseases. Cytopenias, the presence of an ANA titer >1:640, anticentromere positivity, and abnormal nailfold capillaroscopy were associated with evolution to specific diagnostic entities.

Although not well characterized, the presence of certain clinical features may predict evolution into a specific disease [12,62,66-68]:

●In a study of 148 UCTD patients with antibodies to Ro/SSA, for example, 24 percent developed a defined condition, mainly SLE and Sjögren's disease [67]. The specificity of anti-Ro/SSA may predict the evolution of the condition in that those with isolated antibodies to the 60 kD antigen were more likely to develop SLE, while those with antibodies to both 52 and 60 kD antigens correlated with development of Sjögren's disease [68]. (See "The anti-Ro/SSA and anti-La/SSB antigen-antibody systems".)

●In one study of 213 patients with symptoms of early undifferentiated systemic rheumatic disease for at least one year, the development of lupus at 5 years was associated with younger age, African-American ethnicity, and the presence of alopecia, serositis, discoid lesions, a positive Coombs test, a homogeneous ANA pattern, anti-dsDNA and anti-Sm antibodies, and false positive test for syphilis [12]. The strongest predictors were discoid lesions, homogeneous ANA, anti-Sm antibodies, and serositis.

As noted previously, there are some patients who will develop manifestations of more than one discrete systemic rheumatic syndrome. These occur in all possible combinations of rheumatologic/autoimmune symptoms [69]. Some of these conditions occur with sufficient frequency that they have been given their own unofficial designations: RA and SLE is termed "rhupus" and systemic sclerosis and dermatomyositis called "sclerodermatomyositis." Arguably, the most well-defined of these overlap syndromes is mixed connective tissue disease (see "Mixed connective tissue disease"). We and others have emphasized that diagnostic labels have considerable implications and need to be used with caution [3,70,71].

Also, patients with "overlap" rheumatic diseases differ in certain ways from those with single disorders. As an example, in a retrospective analysis with 23 patients with various overlap syndromes, "rhupus" developed sequentially whereas "sclerodermatomyositis" developed simultaneously [72].

A retrospective analysis of the manifestations of patients with overlap syndromes noted that patients with rhupus tended to develop manifestations sequentially, whereas those with sclerodermatomyositis tended to have simultaneous presentation [73]. Patients with rhupus, in addition to clinical and serologic findings consistent with SLE, may have radiographic evidence of erosive arthropathy [15]. The finding of elevated values of anti-CCP antibodies in some patients with rhupus supports the concept that, in some of these patients, this overlap syndrome may represent a true concordance of the two diseases rather than an atypical presentation of the arthropathy of SLE [74]. Given drug-induced lupus can be seen with tumor necrosis factor (TNF) inhibitors, which are used frequently in treatment of RA, discretion is needed in differentiating between a true overlap disorder between SLE and RA versus one that is drug-induced. (See "Drug-induced lupus", section on 'Clinical presentation'.)

Almost without exception, treatment is directed to the particular clinical manifestations and is not predicated on a definitive diagnosis. In one series, for example, MTX was most effective for the treatment of arthritis and dermatitis but had little effect on central nervous system abnormalities or on serositis [75]. An intriguing report asserted the importance of vitamin D supplementation for patients with USRD who were deficient, documenting improvement in immunoregulatory function (Treg/T17 ratio) and suggesting that continuous vitamin D treatment might inhibit the transition of undifferentiated to defined rheumatic diseases [76]. Finally, as some data suggests that hydroxychloroquine may prevent or delay the progression to SLE, this may be an option for a select group of patients [77].

GENERAL APPROACH TO THE PATIENT — The clinical issues which challenge clinicians confronting patients with either undifferentiated rheumatic disease or an overlap syndrome are to determine when, if ever, to render a specific diagnosis, what diagnostic studies to undertake, what management interventions to consider, what outcomes to expect, and how to monitor such patients. Our approach to these patients is summarized in the algorithm (algorithm 1) [13].

Patients presenting without an obvious diagnosis need thoughtful evaluation and examination with attention to the diagnostic entities listed in the table (table 1). There are no consensus guidelines or an evidence-based approach for this [78], and our approach varies depending upon the presenting clinical features; we do not think there is a formulaic approach that will fit all of these patients.

These patients should undergo a thorough physical examination (including fundoscopy) as well as routine diagnostic studies (complete blood counts, differential and platelet counts, complete metabolic panel, urinalysis with microscopic exam). Additional laboratory testing and other studies might include antinuclear antibodies (ANA); anti-double-stranded DNA (dsDNA), -Sm, -RNP, -centromere, -Ro/SSA, -La/SSB, -Jo-1, -Scl-70, and -PM1 antibodies; antineutrophil cytoplasmic antibodies (ANCAs); creatine kinase; a myositis-specific antibody panel; rheumatoid factors; anticyclic citrullinated peptide (CCP) antibodies; thyroid function tests and thyroid peroxidase antibody (TPO); erythrocyte sedimentation rate; C-reactive protein; hepatitis panel; serum protein electrophoresis; a cryoglobulin determination; and nailfold capillaroscopy. For some patients, the possibility of a paraneoplastic syndrome may be in the differential diagnosis and should also be evaluated appropriately. Similarly, certain infectious diseases, particularly viral syndromes, like parvovirus B19, should also be considered. Imaging, biopsy, and other studies will be carried out on certain patients (ie, those presenting with interstitial lung disease or skin lesions, respectively).

While not yet standardized, testing for myositis-specific antibodies when indicated may assist the clinician. Given certain phenotypes of inflammatory myopathies can be amyopathic or hypomyopathic, while others may have more predominant lung, joint, or skin manifestations, the presence of a myositis-specific antibody may help to both diagnose and potentially prognosticate. (See "Overview of and approach to the idiopathic inflammatory myopathies", section on 'Myositis-specific autoantibodies'.)

Follow-up intervals will reflect the nature and severity of symptoms, rapidity of change (if any), and patient and clinician comfort with the diagnostic uncertainty. Patients with new-onset Raynaud phenomenon or new onset of inflammatory arthritis are seen and evaluated monthly for the first three months. We generally see relatively asymptomatic patients at intervals of approximately six months, and symptomatic patients at the same frequencies we would see patients with recognized rheumatic diseases with similar manifestations. We tend to repeat most, if not all, diagnostic studies obtained originally on our follow-up evaluations. We avoid diagnostic labels not clearly supported by established criteria. We tolerate uncertainty as best we can, and help patients and caregivers accept this as well. Patients are managed expectantly and according to their major clinical and symptomatic features (eg, arthritis, glomerulonephritis, Raynaud phenomenon).

SUMMARY AND RECOMMENDATIONS

●Patients with undifferentiated rheumatic diseases and/or overlap syndromes who cannot be definitely diagnosed with a well-characterized systemic rheumatic disease may include up to 25 percent of rheumatic disease patients. These patients often exhibit one of several disease patterns, manifesting multiple nonspecific clinical or serologic abnormalities, sometimes of more than one defined rheumatic disorder (table 1). (See 'Introduction' above.)

●The Raynaud phenomenon (RP) may be an early sign of systemic rheumatic disease in as much as 10 percent of individuals, but one-third of such patients may evolve into an undefined systemic rheumatic disease syndrome (USRD). Patients with RP should be evaluated with a careful history and physical exam, including nailfold capillaroscopy, testing for antinuclear antibodies (ANA), and, in selected cases, testing for the presence of cryoglobulins or other cryoprecipitable proteins. (See 'Raynaud phenomenon' above and 'Nailfold capillaroscopy' above.)

●A heterogeneous group of patients may exhibit symptoms and signs of inflammatory arthritis suggestive but not diagnostic of rheumatoid arthritis (RA), with an elevated erythrocyte sedimentation rate and a positive ANA or rheumatoid factor and with a dramatic response to antiinflammatory therapy. About 10 percent of such patients ultimately fulfill criteria for another systemic rheumatic disease, and 5 to 10 percent will acquire other features of systemic rheumatic disease but still cannot be diagnosed with a specific well-differentiated rheumatic disorder. (See 'Polyarthritis' above.)

●Interstitial lung disease (ILD) may precede or occur at or after onset of various connective tissue and rheumatic disorders. A histologic pattern referred to as nonspecific interstitial pneumonia pattern (NSIP) is commonly found in lung biopsy specimens of patients with ILD associated with undifferentiated connective tissue diseases (UCTD), systemic lupus erythematosus (SLE), Sjögren's disease, systemic sclerosis, and dermatomyositis. The search for an underlying CTD in patients with ILD is warranted due to the generally more favorable prognosis compared with idiopathic disease. Myositis-specific antibodies may be useful in such a situation. (See 'Nonspecific interstitial pneumonia' above.)

●Patients who present with undifferentiated systemic rheumatic (connective tissue) disease or who have overlapping clinical or serologic features of two or more diseases have been estimated to comprise 15 to 25 percent of tertiary rheumatology referrals. A premature diagnosis of one of the discrete rheumatic diseases should be avoided, since the undifferentiated nature of the syndrome usually persists. Among those destined to develop a defined connective tissue disease, which is more common than resolution of the syndrome, evolution generally occurs within two to five years. Treatment is generally directed to the particular clinical manifestations. (See 'Early undifferentiated systemic rheumatic disease' above.)

●Patients either with undifferentiated rheumatic disease or with an overlap syndrome should undergo a careful diagnostic evaluation (algorithm 1) which considers the most likely diagnostic entities (table 1). Studies that should be obtained depend upon the presenting clinical features. A specific diagnosis should be made only when patients fulfill appropriate classification criteria; other patients should be followed and reassessed periodically and should be managed according to their major rheumatologic symptom(s). (See 'General approach to the patient' above.)