Past Medical History, Medication Use, and Exposures to Toxins or Infectious Diseases
It is critical to know the full medical history of any patient suspected of having vasculitis. Other diseases may either have vasculitis as a component of the illness (e.g., lupus) or may cause skin lesions that mimic vasculitis. Drug-induced vasculitis (DIV) is common and skin lesions, usually but not always purpura, are the most common manifestation of DIV.9 The list of drugs reported to cause vasculitis is enormous with almost every class of medication implicated in possible cases of DIV. It is useful to ask about prescription, nonprescription, and “alternative” or herbal mediation use in the prior 6–12 months since the effect of some medications may persist after usage ends. Patients should also be asked about use of illegal or recreational drugs since several such agents, including methamphetamines, cocaine, and others have been implicated in cases of vasculitis. Occupational or other exposure to nondrug toxins should be asked about.
The patient should be asked about not only usual signs and symptoms of infection, but also recent travel, contacts with sick individuals, and risks for sexually transmitted diseases.
Beyond a careful and full assessment of the skin, a multisystem examination is useful to determine whether symptoms are associated with objective abnormalities, or whether there are findings that a patient has not noticed. Vital signs are essential, but a patient with normal blood pressure can still have severe glomerulonephritis. The eyes should be inspected for redness and proptosis. The anterior nasal cavity can be easily visualized with an otoscope. Evidence of lymphadenopathy should be sought. Cardiac, lung, and abdominal examinations can give clues to underlying disease, but normal examinations do not rule out pathology. Similarly, absent pulses, asymmetric blood pressure readings, and bruits are helpful but imperfect measures to screen for large-vessel vasculitis. A complete joint examination is important and any findings suggestive of synovitis (joint swelling, warmth, redness) must be further investigated; however, many patients with vasculitis will have arthralgias without joint effusions. A full neurological examination is one of the most valuable components of the evaluation and triage of a patient suspected of having vasculitis; subtle sensory and even motor abnormalities may be missed on initial evaluation.
The more detailed and expert examinations that can be performed by ophthalmologists and otolaryngologists are often extremely helpful in evaluating patients suspected of having vasculitis. Urgent referral is often indicated in patients with concerning symptoms such as new visual impairment, painful or red eyes, hoarseness or stridor, or hearing loss.
Diagnostic Testing for Suspected Vasculitis
Given the broad range of entities that fall under the category “vasculitis”, and the even larger number of diseases that are also reasonably considered when evaluating a patient suspected of having vasculitis, a vast number and range of diagnostic tests are often considered in such cases. However, obtaining a thorough medical history and conducting a detailed physical examination should enable the clinician to limit the types of vasculitis under consideration and prioritize the ordering of diagnostic tests. Not all tests need be ordered for all patients suspected of having vasculitis. This approach must of course be balanced by the possibility of atypical presentations of vasculitis as well as a range of infections, malignancies, and other diseases in the differential diagnosis of such patients. Evaluation for possible vasculitis usually occurs in parallel to evaluation for other processes.
The method for diagnosing vasculitis depends on the type of vasculitis suspected, which is often based on the size of vessel involved. Vasculitides affecting the skin usually involve small- and medium-sized vessels and these vessels are amenable to biopsy (Fig. 164-3). Given the ease and low risk of skin biopsies, they play an important role in diagnosing vasculitis, and an equally important role in establishing a diagnosis other than vasculitis. A standard punch biopsy is sufficient to diagnose small-vessel vasculitis but a deeper and wider excision may be necessary to capture information on medium-sized vessels.10,11 Lesions that should be approached with a deeper biopsy include subcutaneous nodules, livedo reticularis, or deep ulcers (Fig. 164-4). Many types of small-vessel vasculitis may also involve medium-sized skin vessels. It is important to realize that the difference between “small” and “medium” vessels is somewhat subjective and skin pathologists make such distinctions more often than might other pathologists who see larger biopsy specimens.
Different skin manifestations of systemic vasculitis. A. Purpura. B. Bullae. C. Ulcer.
A. A patient with polyarteritis nodosa with “starburst” livedo made up of a cluster of nodular lesions. B. Histopathology of skin lesions in polyarteritis nodosa showing segmental necrotizing arteritis.
Sometimes the presence of a typical clinical syndrome makes biopsy unnecessary. For example, Henoch–Schöenlein purpura in children is often diagnosed on clinical grounds alone and some cases of ANCA-associated or cryoglobulinemic vasculitis can be diagnosed confidently by combining clinical features with specific serologic tests. Behçet disease and Kawasaki disease are diagnosed based on the clinical syndromes; biopsy is usually not performed on the skin lesions that are common in these diseases, and such biopsies are often nondiagnostic.
It is generally recommended to biopsy a skin lesion that has been clinically apparent for less than 48 hours, if possible, to maximize the chance of finding the typical features of acute neutrophilic vasculitis, including fibrinoid necrosis, extravasation of erythrocytes, extravasation of neutrophils with release of nuclear debris (leukocytoclasia), and the presence of immune deposits.10,11 Processing of tissue is different for conventional histopathology or immunofluorescence testing; if immunofluorescence is desired, then either two specimens need to be obtained, or a single specimen needs to be divided before processing. The latter approach may however damage the tissue.10
As discussed throughout this chapter, the histologic finding of leukocytoclastic vasculitis is helpful in confirming the diagnosis of vasculitis but does nothing to establish an etiology from among the broad number of possibilities. Microscopy sometimes reveals features that are suggestive but not diagnostic of vasculitis, such as leukocytoclasia without fibrinoid necrosis. The finding of a perivascular infiltrate, particularly if it consists predominantly of mononuclear cells but even if it is neutrophilic, is also nonspecific. Certain features, when seen in addition to leukocytoclastic vasculitis, are strongly suggestive of particular diseases, such as extravascular granulomas with geographic necrosis (GPA), or eosinophil-rich extravascular granulomas (Churg–Strauss syndrome; CSS),12 but these features are seen in a minority of biopsies in these diseases.
A predominance of IgA over IgG/IgM by immunofluorescence is suggestive but not diagnostic of Henoch–Schönlein purpura. The presence of deposits of IgG, IgM, and/or complement is suggestive of one of several immune-complex-mediated etiologies, including drug hypersensitivity, postinfectious vasculitis, cryoglobulinemia, and vasculitis secondary to systemic lupus erythematosus, Sjögren syndrome, or rheumatoid arthritis.12
Vasculitis is often diagnosed by biopsy of other organs, such as kidney, lung, muscle, or peripheral nerve or even from surgical specimens (Fig. 164-5). Kidney or lung biopsies are more likely than skin biopsies to show pathology diagnostic of a particular disease. Nonetheless, a skin biopsy establishing the diagnosis of vasculitis may preclude the need for more invasive biopsies.
A. Lung histopathology from a patient with granulomatosis with polyangiitis (Wegener's) demonstrating necrosis, giant cells, and mixed cellular inflammation. B. “Geographic necrosis” in a low-power view of an open lung biopsy specimen from a patient with granulomatosis with polyangiitis (Wegener's).
Although individual laboratory tests on their own are almost never diagnostic for vasculitis, such tests are essential in the evaluation of a patient in whom cutaneous vasculitis is being considered. Laboratory testing may identify organ systems involved in the disease process, especially renal disease. Furthermore, in the proper setting, selected serologic tests may establish an etiology for vasculitis. However, serologic tests usually complement rather than substitute for biopsy, particularly in a patient with skin lesions that can be readily biopsied.
Tests for renal disease are the most important laboratory tests to order in evaluating a patient suspected of having vasculitis since renal disease is common in many vasculitides and is rarely accompanied by signs or symptoms until end-stage renal failure occurs. Urinalysis, including both dipstick and microscopic examinations, should be performed on all patients in whom vasculitis is suspected, and repeatedly in patients in whom vasculitis of small- or medium-sized vessels is established in another organ system. The presence of any blood on the routine dipstick tests needs to be followed by an examination for red blood cell casts by someone specifically trained to look for casts (many nephrologists, some rheumatologists, but few laboratory technicians in North America). Measurement of serum creatinine is critical to estimate the glomerular filtration rate (GFR). Small changes in creatinine, even within the normal range, may be early evidence of decline in GFR. Although small-vessel vasculitis affecting the glomeruli is expected to produce hematuria, usually accompanied by red blood cell casts and proteinuria, vasculitis affecting only medium-sized vessels (e.g., polyarteritis nodosa) typically produces either isolated hematuria or a normal urinalysis. Urinalysis and serum creatinine are equally important tests and are complementary; neither alone is sufficient to exclude renal disease in vasculitis.
Vasculitis, particularly polyarteritis nodosa, can involve the liver, but significant hepatic dysfunction is rare. Liver function tests are thus of limited value in diagnosing vasculitis, but they do provide a baseline against which future values can be compared if, as is often the case, potentially hepatotoxic drugs are to be used for treatment. Liver function tests can also provide an early hint at infection with hepatitis B or C viruses, both of which are associated with vasculitis, but do not substitute for serologic testing for these infections. Normal liver function tests do not rule out infectious hepatitis.
A complete blood count should be ordered on all patients suspected of having vasculitis. Many patients with active vasculitis have anemia and/or thrombocytosis, but the same is true of a wide range of inflammatory diseases. Severe anemia can be a clue to serious gastrointestinal involvement from various forms of vasculitis. The white blood cell count and differential can also be clues to the presence of infection or hematologic malignancy. However, leucocytosis is usually nonspecific and is also commonly caused by use of glucocorticoids. An elevated absolute eosinophil count is found in most untreated patients with CSS and a count greater than 1,000 cells/μL helps differentiate this disease from asthma and atopy.
The erythrocyte sedimentation rate (ESR) and levels of C-reactive protein (CRP) are elevated in many patients with vasculitis, but the diagnostic sensitivity and specificity of these tests are not particularly high and thus these tests are not particularly helpful in either establishing or excluding a diagnosis of vasculitis. Furthermore, the levels of ESR and CRP do not correlate well with stage or severity of disease. ESR and CRP are often elevated in conditions that mimic vasculitis in the skin, as well as in many serious systemic diseases, including infections and malignancies. Patients with active vasculitis can have normal ESR and CRP values and patients may remain in clinical remission despite persistent elevation of these markers after treatment.
Testing for autoantibodies is often a critical component of establishing the type of vasculitis present but it is important to recognize that serologic testing on its own is never diagnostic and should never substitute for clinical judgment.
Testing for ANCA and antiglomerular basement membrane (anti-GBM) antibodies, as well as antinuclear antibodies (ANA) to address the alternative possibility of systemic lupus erythematosus, is advised for any patient presenting with pulmonary hemorrhage and/or acute renal insufficiency with an active urinary sediment. ANCA-associated vasculitis and lupus can present with vasculitis of the skin, but anti-GBM disease does not, so the latter topic will not be discussed further.
Antineutrophil Cytoplasmic Antibodies
Approximately 90% of patients with microscopic polyangiitis, 75% of patients with GPA, and 40% of patients with CSS will test positive for ANCA.13–16 Modern ANCA testing includes both immunofluorescence staining of neutrophils for the cytoplasmic (c-ANCA) or perinuclear (p-ANCA) patterns and ELISAs for specific autoantigens [proteinase-3 (PR3) and myeloperoxidase (MPO)].16–18 Specificity of positive testing for anti-PR3 and anti-MPO antibodies for ANCA-associated vasculitis is quite high,17–19 but specificity of p-ANCA staining in the absence of anti-MPO antibodies is low. Thus, positive tests for ANCA by ELISA are essential to consider ANCA testing positive for purposes of diagnosing vasculitis.
The predictive value of positive ANCA testing depends on the setting. In cases of biopsy-proven vasculitis or clinical “surrogates” of a biopsy of vasculitis, such as diffuse alveolar hemorrhage or acute renal failure with an “active” urinary sediment, positive testing for anti-PR3/MPO ANCA is highly specific. In the setting of nonspecific constitutional and musculoskeletal symptoms, the positive predictive value of ANCA testing is lower.
Testing for ANA and related autoantibodies is useful when there is suspicion of systemic lupus or Sjögren syndrome. ANA testing is extremely sensitive (>95%) but not specific for the diagnosis of lupus. With the exception of anti-Ro (SSA) antibodies, additional tests for specific nuclear antigens, including double-stranded DNA, Smith, RNP, and La (SSB), should only be ordered if the ANA is positive and lupus is still under consideration. Only 80% of patients with Sjögren syndrome test positive for rheumatoid factor (RF), anti-Ro (SSA), or anti-La (SSB) antibodies, so negative tests do not rule out this diagnosis.
Testing for rheumatoid factor is rarely useful in establishing either the diagnosis or specific type of vasculitis. The sensitivity and specificity of rheumatoid factor for Sjögren syndrome or cryoglobulinemic vasculitis are low. While at least 70% of patients with rheumatoid arthritis test positive for rheumatoid factor, the test is positive in more than 95% of patients with rheumatoid vasculitis.20 However, since rheumatoid vasculitis typically occurs in patients with longstanding, severe rheumatoid arthritis, such testing has little additive value.
Paraproteins (Abnormal Immunoglobulins, Including Cryoglobulins)
Cryoglobulins are immune complexes (immunoglobulins and their target antigens) that precipitate in the cold and are associated with clinical syndromes in which vasculitis is a prominent component (see Chapter 169). Cryoglobulinemia most commonly results from chronic infection with hepatitis C virus, but rheumatoid arthritis, systemic lupus erythematosus, Sjögren syndrome, and hematologic malignancies are also all associated with cryoglobulinemia. Testing for cryoglobulins requires careful attention to specimen handling and processing since incorrect practice at any one of several steps results in a high false-negative rate. Similarly, standard serum protein electrophoresis testing may not pick up some immunoglobulin clones, and immunofixation electrophoresis is a more comprehensive screen for clonal immunoglobulins.
Vasculitis has also been associated with monoclonal gammopathies (myeloma, plasmacytoma, or lymphoma) in the absence of cryoglobulinemia.21
Total hemolytic complement is measured using the CH50, but since this assay is cumbersome and suffers from variability between laboratories, measurement in serum of the complement proteins C3 and C4 is usually sufficient, and is useful for assessing patients with cutaneous vasculitis in several settings. In patients with cryoglobulinemic vasculitis C4 levels are usually severely depleted whereas C3 levels are less depleted or even normal.22,23 One or both of these components are low in 70% of patients with rheumatoid vasculitis,20 which is useful because rheumatoid arthritis is generally not associated with low circulating complement. In contrast, because systemic lupus erythematosus is commonly associated with low complement in a variety of settings, low complement helps raise the suspicion for lupus but is not specific for vasculitis in SLE. A subset of patients whose cutaneous vasculitis presents as urticaria (hence the term “urticarial vasculitis”), but who cannot be diagnosed with lupus or another underlying disease, have depletion of complement (see Chapter 163).
Selected Testing for Infectious Diseases
Many infections can cause skin lesions that either include vasculitis or mimic vasculitis. Chronic infection with hepatitis C virus is strongly associated with cryoglobulinemic vasculitis, and it can also be associated with polyarteritis nodosa in the absence of cryoglobulins.24 Chronic hepatitis B virus infection was the cause of many cases of polyarteritis nodosa prior to the widespread adoption of vaccination programs.25 Thus, patients with known or suspected vasculitis affecting small- or medium-sized arteries should be screened for hepatitis B and C infections.
Endocarditis can cause both true vasculitis, presumably through deposition of immune complexes, and lesions that mimic vasculitis, through septic emboli. Blood cultures are therefore appropriate to order for some patients suspected of small-vessel vasculitis.
Numerous and diverse infections have been implicated in causing secondary vasculitis, usually of small vessels and limited to the skin. Testing for specific organisms should therefore be based on a history of exposure or a suspicious clinical syndrome (e.g., sore throat or acute diarrhea).
Several uncommon infections directly infect and damage vascular endothelial cells and thus produce lesions that can either be regarded as vasculitis or as mimics of vasculitis; numerous organisms have been implicated, mostly in the form of case reports.
Screening for Drugs of Abuse
Toxicology screens for commonly used drugs of abuse may be appropriate for some clinical situations where vasculitis is suspected. In particular, both cocaine and methamphetamines have been associated with cutaneous vasculitis and/or arterial vasospasm, and nasal inhalation of cocaine can produce destructive nasal disease as severe as that seen in ANCA-associated vasculitis, although certain clinical features may help distinguish the two causes.26,27
Diagnostic Imaging Studies
A chest radiograph is an appropriate screening test for any patient suspected of having vasculitis. For a patient with pulmonary symptoms, computed tomography (CT) is usually indicated, since plain radiographs will frequently not detect small nodules or subtle but significant infiltrates. In patients diagnosed with GPA, microscopic polyangiitis, or CSS, a screening CT is indicated for staging purposes and to establish a baseline, even in asymptomatic patients.
If subglottic stenosis is suspected, CT of the neck/trachea can be a helpful adjunct to direct laryngoscopy.
Sinus involvement is extremely common in GPA and CSS, and the ability to evaluate the sinuses on physical examination is limited even for an otolaryngologist. CT of the sinuses can help assess the possibility of GPA or CSS and is useful in staging and restaging disease once one of those diagnoses is made and treatment is initiated. However, the CT appearance of sinus inflammation in these diseases does not allow discrimination from other causes of sinusitis, and patients with prior damage from vasculitis often have persistent abnormalities. Nasal inflammation may be better assessed by physical exam than by CT.
Angiography has a central role in the diagnosis and management of large- and medium-vessel vasculitis. Conventional catheter-based dye angiography has the highest resolution but is an invasive procedure and still does not allow visualization of most small vessels. Angiography based on CT and magnetic resonance (MR) is increasingly replacing the use of catheter-based angiography in large-vessel vasculitis (GCA and Takayasu's).28
The role of angiography in the diagnosis of vasculitis of the skin is limited to either establishing the underlying type of vasculitis (e.g., abdominal angiography demonstrating multiple aneurysms and stenoses in polyarteritis nodosa) or evaluating the arterial supply in patients with gangrene.
Nerve Conduction Studies and Electromyography
Nerve conduction studies should never replace a full neurologic examination and most patients with neurologic manifestations of vasculitis do not need such testing. Thus, nerve conduction testing is not recommended for screening asymptomatic patients, but can be useful for providing objective evidence of neuropathy and for distinguishing between compressive (i.e., mechanical) and noncompressive neuropathy, with the latter type including neuropathy due to vasculitis and many other medical causes. Electromyography (EMG) can establish the presence of myopathy, but usually not the cause. Nerve conduction studies are painful and require expertise not always readily available.
An audiogram is critical in diagnosing and distinguishing between conductive and/or sensorineural hearing loss. Hearing loss is a commonly missed manifestation of small-vessel vasculitis, including among elderly patients.29 Sensorineural hearing loss is a cranial neuropathy and may rapidly lead to irreversible hearing loss. Although an audiogram is not generally indicated for screening an asymptomatic patient, a baseline audiogram is advised for all patients with an established diagnosis of ANCA-associated vasculitis (GPA, microscopic polyangiitis, or CSS).