E.William St. Clair
The major vasculitis syndromes encompass a broad spectrum of clinical disorders (see Table 27-1 for a partial listing of the recognized syndromes). 1, 2  These diagnostic categories have many overlapping clinical, laboratory, and histopathologic features, a fact that has created much nosologic confusion and impeded the acceptance of a standard nomenclature.  Disease definitions have been updated to reflect our increased knowledge of systemic vasculitis.  A consensus conference held in Chapel Hill, North Carolina (CHCC, or Chapel Hill Consensus Conference) partially revised the current scheme;  of most note it created a subgroup called microscopic polyangiitis.3  In addition, the American College of Rheumatology (ACR) has developed criteria for classifying patients with systemic vasculitis in clinical studies.4
The heterogeneity of the systemic vasculitides not only poses diagnostic challenges but also shapes management.  Appropriate treatment of these complex diseases requires a comprehensive assessment of each patient in order to determine the extent and severity of organ system involvement and to exclude toher diagnoses that may be confused with vasculitis (e.g., infection, neoplasm).  Continual reevaluation of disease activity and close monitoring for possible treatment-related complications are essential for attaining the best possible outcomes.
This chapter reviews the treatment of systemic vasculitis.  It aims to characterize the natural history of each of the disorders, to critically review the clinical studies, and to provide guidelines for their management on the basis of expert opinion and the best clinical evidence.  The specific types of vasculitis discussed in this chapter are PAN, microscopic polyangiitis (MPA), Churg-Strauss syndrome (CSS), Wegener’s granulomatosis (WG), essential type II mixed cryoglobulinemia, Henoch-Schönlein purpura (HSP), hypersensitivity vasculitis, and Cogan’s syndrome.
The pathologic hallmark of PAN is a small or medium-sized necrotizing arteritis.  Its clinical features include fever, weight loss, cutaneous ulcers, livedo reticularis, muscle pain and weakness, arthralgias, and arthritis, neuropathy, abdominal pain, ischemic bowel, testicular pain or tenderness, hepertension, and renal failure.  Although PAN classically spares the lung, a polyangiitis overlap syndrome with PAN-like features and pulmonary involvement has been described.5  Many of these overlap cases fit current definitions of CSS and MPA.
A focal segmental necrotizing glomerulonephritis without interlobar or intralobar artery involvement has been termed microscopic polyarteritis.6 It is often accompanied by a small or medium-sized vessel vasculitis in other organ systems.6 Microscopic polyarteritis, originally a pathologic designation, has been renamed microscopic polyangiitis (MPA), a necrotizing, pauciimmune vasculitis affecting small vessels (capillaries, venules, or arterioles).3  By definition, MPA may also target small and medium-sized arteries.  The dominant clinical and pathologic features of MOA are glomerulonephritis and pulmonary capillaritis with alveolar hemorrhage.  Other clinical features of MPA are similar to those of classic PAN except that PAN entails a lower frequency of peripheral neuropathy.7,8   Although angiography has not been performed uniformly in patients with MPA, abnormal angiograms with microaneurysm and vessel stenosis have been only rarely observed in this disease subgroup.7  Unlike PAN, MPA has not been associated with hepatitis B virus infection.  Serologically, serum from about 60% of patients with MPA contain antneutrophil cytoplasmic antibodies (ANCA) with a perinuclear staining pattern. 7, 8 More than 90% of these sera are positive for reactivity with myeloperoxidase (MPO).7 
Classic PAN was redefined at the CHCC as a necrotizing vasculitis of medium-sized or small arteries without glomerulonephritis and without vasculitis in arterioles, capillaries, or venules.  Most authorities believe that this definition is too restrictive because small-vessel involvement occurs commonly in patients with classic PAN.9  No cases of classic PAN as defined by the CHCC were identified in a 6-year epidemiologic study of systemic vasculitis in Norwich, United Kingdom, in comparison with 13 patients who met criteria for MPA. 10  Of the 13 patients with MPA, 8 also met ACR criteria for PAN.  Thus, the CHCC definition of classic PAN should be revised to include patients with small-vessel involvement whose disease otherwise fits the clinical picture. Patients with classic PAN are characterized by renal vasculitis with hypertension, the absence of glomerulonephritis, abnormal angiograms, a high prevalence of multiple mononeuropathies, and a lack of lung involvement.8  Classic PAN is further distinguished from MPA by its association with hepatitis B infection.11,  12  and the usual absence of serum ANCA.  Not to be confused with microscopic polyarteritis or MPA, cutaneous PAN is a clinical variant of necrotizing vasculitis that predominately targets the skin and subcutaneous tissue and does not usually progress to systemic involvement.
Separating PAN into two subsets, classic PAN and MPA, may provide useful prognostic information.  Guillevin and Lhote suggested that classic PAN is a  monophasic disease with relatively few relapses, implying that treatment of this subset may be limited to a defined time period (e.g., 1 year).9 In comparison, MPA is characterized by a high probability of relapse and often necessitates more prolonged therapy to control the intermittent flares of disease.  The outcome of MPA appears to be similar to that of the other systemic necrotizing vasculitides.7 
CSS, also called allergic angiitis and granulomatosis, is characterized by asthma, pulmonary vasculitis, and peak eosisnophil counts exceeding 1.5 x 109/L.13, 14  The heart, nervous system, gastrointestinal and urinary tracts, and joints as well as the lungs are the principal targets of this vasculitis.  A minority of affected patients exhibit severe renal disease.  CSS musst be differentiated from other clinical disorders associated with marked eosinophilia, such as eosinophilic leukemia, hypereosinophilic syndrome, Löffler’s syndrome, chronic eosinophilic pneumonia, eosinophilic gastroenteritis, parasitic infections, and bronchopulmonary aspergillosis.
Laboratory studies in patients with PAN, MPA, and CSS reflect the heightened inflammatory state (e.g., increased levels of acute-phase reactants) and ischemic injury to affected organ systems.  These findings are nonspecific.  ANCA are associated with certain forms of vasculitis and may provide clues to the diagnosis.  Finding the typical arteriographic abnormalities15 and histopathologic changes of necrotizing vasculitis16 remains the gold standards for the diagnosis of vasculitis.
TABLE 27-1
Classification of the Vasculitic Syndromes
Polyarteritis Group
Classic polyarteritis nodosa Medium- and small-vessel vasculitis
Microscopic polyangiitis Necrotizing vasculitis of small and medium-sized arteries
Necrotizing vasculitis affecting capillaries, venules, and arterioles
Churg-Strauss syndrome Eosinophilic and granulomatous inflammation of the respiratory 
 tract and necrotizing vasculitis affecting small and medium-sized vessels
Polyangiitis overlap syndrome Overlapping features of specific types of vasculitis
Wegner’s Granulomatosis
Granulomatous vasculitis affecting small to medium-sized vessels
Small-Vessel Vasculitis                        Inflammation affecting capillaries, venules, or arterioles
Cutaneous leukocytoclastic vasculitis Small-vessel vasculitis limited to the skin (no systemic involvement)
Essential mixed cryoglobulinemia Small-vessel vasculitis with serum cryoglobulins (usually skin and glomerulli)
Henoch-Schönlein purpura     Small-vessel vasculitis with IgA-dominant immune deposits (usually skin, gut, and                                       glomeruli) and arthritis     
Serum sickness Small-vessel vasculitis resulting from repeated injections of a foreign antigen
Vasculitis associated with 
   malignancies       Small-vessel vasculitis associated with cancer (hairy cell leukemia, lymphoma, solid tumors)
Vasculitis associated with other
   Primary disorders Small-vessel vasculitis associated with rheumatoid arthritis, systemic lupus erythematosus, other connective tissue diseases
Primary Angiitis of the Central
   Nervous System Vasculitis Small-and medium-vessel vasculitis of the central nervous system
Kawasaki’s Disease Small-, medium-, and large-vessel arteritis associated with mucocutaneous lymp node syndrome (usually occurs in children and commonly results in coronary artery involvement)
Giant Cell Arteritides Granulomatous inflammation of the large vessels
Temporal arteritis Granulomatous inflammation of the extracranial branches of the carotid artery ( may also   involve the aorta and its major branches); occurs in patients more than 50 years of age
Takayasu’s arteritis Granulomatous inflammation of the aorta and its major branches; occurs in patients less than 50 years of age
Miscellaneous Vasculitides Medium- and large-vessel vasculitis
    (Cogan’s Syndrome, Behçet’s Small-, medium-, and large-vessel vasculitis
      IgA = immunoglobulin A.
Natural History of Disease
Our understanding of the natural history of PAN (and MPA) and CSS derives from clinical observations made before the introduction of glucocorticoids.  These clinical and histopathologic descriptions did not take into account the major subdividions of systemic vasculitis or distinguish classic PAN from MPA.  Instead, PAN, CSS, and other vasculitic syndromes were grouped into the broader category of “periarteritis nodosa.”  Patients with periarteritis nodosa who were seen between 1946 and 1962 at the Mayo Clinic and left untreated had 1- and 5-year survival rates of 35% and 13%, respectively.17  Initial presentation with renal or gastrointestinal disease was associated with a worse outcome; most early deaths were caused by active vasculitis.
Before the introduction of antiviral therapy, outcomes of patients with PAN were not substantially worsened by concomitant hepatitis B virus (HBV) infection.  In one study, the 5-year survival rates among patients with systemic necrotizing vasculitis with and without HBV markers were 61% and 58%, respectively.18  However, the HBV-positive group had a significantly higher mortality rate (30%) during the first year of disease than did the HBV-negative group (15%).18   The most common causes of death in the patients with PAN and HBV infection were gastrointestinal bleeding and bowel perforation.18  
Patients with CSS seemingly fare no better without therapy than does the periarteritis nodosa group as a whole.  A retrospective analysis showed that untreated patients with CSS have a 1-year survival rate of 4%.19  Allergic rhinitis and asthma, the prodrome to CSS, were found to precede the onset of vasculitis by as long as 7 years.14  Deaths in patients with CSS have commonly resulted from pulmonary compromise, cardiac failure, renal failure, and cerebral hemorrhage.14, 19 
Uncontrolled Treatment Studies
Glucocorticoids became widely accepted as efficacious therapy for systemic necrotizing vasculitis on the basis of uncontrolled data from retrospective studies.  Patients with PAN who were treated with glucocorticoids had survival rates of 60% to 71% at 1 year and 48% to 55% at 5 years.17,20  Further improvement in the prognosis of PAN and CSS came with the use of cytotoxic therapy.  Fauci and associates21  reported in 1979 that treatment with 2 mg/kg/day of oral cyclophosphamide produced a high rate of remissions in 17 patients with PAN in whom disease had been previously refractory to glucocorticoid therapy.  Although 3 of the patients in this study died, the other 14 patients achieved a complete or partial remission and within a few months were able to tolerate a reduction in glucocorticoid doses to a less toxic, alternate-day regimen.  Because the patients in this study had had disease of more than 2 years’ duration when cytotoxic therapy was initiated, the outcomes observed during followup may have been affected by the natural history of disease and previous glucocorticoid therapy as well as by the addition of cytotoxic agents.
The experience at the Mayo Clinic, however, did not yield improved outcomes with cytotoxic therapy.20  In a retrospective study, Nachman and colleagues22  assessed treatment outcomes in 107 patients with ANCA-associated MPA and necrotizing, crescentic glomerulonephritis.  None of the patients in that study had WG.  All of the patients initially received 7 mg/kg/day of methylprednisolone for the first 3 days, which was followed by nonrandom assignment to one of three treatment regimes:  prednisone (1 mg/kg/day); prednisone (1 mg/kg/day) plus oral cyclophosphamide (2 mg/kg/day); or predisone (1 mg/kg/day) plus 6 monthly intravenous pulses of cyclophosphamide (0.5 to 1.0 mg/m2).  The remission rate was significantly higher in the two cyclophosphamide-treated groups (84%) than among the patients receiving glucocorticoids alone (56%).  Much of the mortality in these retrospective studies occurred in the first 2 years of disease.  The results of these studies must be interpreted with caution because nonrandom allocation of patients at various stages of disease into different treatment groups may bias the survival analysis.  For example, a bias against glucocorticoid therapy would occur if the most severely affected patients treated initially with glucocorticoids alone died before they had an opportunity to receive cytotoxi therapy.
The treatment of hepatitis B virus (HBV)-related PAN is complicated by the fact that immunosuppressive therapy may interfere with the body’s immune response to the virus and actually worsen disease outcomes.  The exacerbating effects of immunosuppressive therapy are reflected in the negative impact of glucocortoicoid therapy on the survival of patients with HBV-associated liver disease.  Also, sudden cessation of immunosuppressive therapy may reconstitute cellular immunity and provoke fulminant hepatitis.23
The treatment of HBV-related PAN has evolved with advances in antiviral therapy. The most widely studied agents for HBV-related PAN are vidarabine and interferon-alpha (IFN-alpha).  A retrospective analysis of 66 patients with HBV-related PAN monitored for a mean of 50 + 46 months showed that treatment with glucocorticoids, plasma exchange, andcyclophosphamide was associated with a higher mortality rate (46%) than was treatment with vidarabine, plasma exchange, and a short initial course of prednisone therapy (18%).24   In another series, a mortality rate was 19.5% among 41 patients with HBV-related PAN who had been monitored for a mean of 70 + 45 months after receiving either vidarabine or IFN-alpha coupled with a short course of prednisone and plasma exchange to achieve initial disease control.25 These results suggest that immunosuppressive therapy should be minimized in HBV-positive patients.
Glucocorticoids and cytotoxic agents have also been employed in the treatment of CSS. Patients with CSS who received 40 to 60 mg of prednisone or its equivalent had 1- and 5-year survival rates of 90% and 62%, respectively.26  Most deaths in this group resulted from myocardial infarction and cardiac failure.  Clinical disease inresponsive to glucocorticoids could often be brought under control with azathiiprine26 or cyclophosphamide.26
Prospective Clinical Trial
The French Vasculitis Study Group has conducted several prospective treatment trials in systemic vasculitis.  In their first trial, 71 patients with PAN or CSS were randomly assigned to receive initial treatment either with prednisone plus plasma exchange or with prednisone, plasma exchange, and oral cyclophosphamide.27  Both treatment groups achieved similar remission and survival rates, although the group receiving oral cyclophosamide suffered fewer relapses during follow-up.  Only two patients receiving prednisone plus plasma exchange deteriorated during the first 6 months of treatment. Age over 55 years and an elevated serum creatinine level at presentation were predictive of a worse outcome.  Although initiation of therapy with oral cyclophosphamide did not enchance overall survival in patients with PAN and CSS, this approach was still favored by the trial’s investigators because it prolonged the disease-free interval.
Plasma exchange has been used as initial therapy in systemic vasculitis to afford rapid disease control.  The incremental therapeutic value of plasma exchange was examined in a clinical trial in which patients with PAN can CSS were randomly assigned to receive either prednisone plus plasma exchange or prednisone alone.  The two treatments produced nearly identical remission rates (83%) and 7-year survival rates (79%).28 The potencial efficacy of plasma exchange was further tested in a randomized, controlled trial of 62 patients with severe PAN and CSS.  To be eligible, patients with PAN and CSS had to have at least one factor indicative of a poor prognosis (e.g., renal disease, gastrointestinal tract involvement, cardiomyopathy, central nervous system involvement, loss of > 10% of body weight, or age o >50 years). Eligible patients were randomly assigned to receive plasma exchange or placebo and then treated with glucocorticoids and intravenous (IV) pulse cyclophosphamide.29  Plasma exchange was associated with only three complications:  local hematoma, trnsient hypotension, and sepsis resulting from catheter infection with Staphylococcus.  The trial was stopped after 5 years because the two treatment arms showed similar 5-year cumulative survival rates.  Together, these results argue that plasma exchange should not be routinely prescribed as initial treatment for systemic vasculitis.
The toxic effects of long-term treatment with oral cyclophosphamide has caused investigators to evaluate other immunosuppressive regimens with potentially less toxicity.  Monthly IV cyclophosphamide therapy is one example of a potentially less toxic approach.  In a small, randomized trial of 25 patients with PAN or CSS, 1 year of treatment with oral cyclophosphamide, 2 mg/kg/day, was compared with 1 year of monthly IV cyclophosphamide (0.6 mg/m2 ) pulses.30  There were no significant differences in clinical efficacy between the two regimens.30  However, the frequency of side effects was significantly higher with oral cyclophosphamide. Similar results were observed in another small, randomized trial comparing outcomes of intermittent pulse and daily oral cyclophosphamide therapy in 47 patients with WG and MPA.31   The small sample sizes in these studies preclude any conclusions about the possible equivalence of daily oral and IV monthly cyclophosphamide therapy despite the similarity in response rates.
The concept of induction and maintenance therapy was examined by Adu and associates32  in a 1-year randomized trial of 54 patients with PAN, CSS, and WG.  In one of the treatment arms, daily oral cyclphosphamide served as induction therapy and was followed by a maintenance regimen of azathioprine.  In the other arm, repeated cyclophosphamide pulses were used to first induce and then maintain remission.  There were no significant differences between the two groups in survival or frequency of relapses.  The rate of toxic effects was slightly higher in the patients receiving the daily oral dosage than in those receiving intermittent pulse cyclophosphamide, but this difference was not statistically significant.
These prospective clinical trials are informative because they provide a contemporary prognostic view of PAN and CSS.  The 7- and 10-year survival rates of patients treated with glucocorticoids alone in two of these trials were 72% and 83%, 27, 28 exceeding the 50% long-term survival rates of glucocorticoid-treated patients of prior studies.  Earlier diagnosis, differences in case severity, improved use of glucocorticoids, and closer monitoring of patients may explain the superior outcomes in these prospective trials.  These survival data also reveal that large sample sizes are necessary for a clinical trial to show that oral cyclophosphamide or another treatment strategy is superior to glucocorticoids alone.  None of the clinical trials involving PAN and CSS had sufficient numbers of subjects to yield any definite conclusions about the superiority of a particular treatment strategy over glucocorticoids alone with survival rate as the primary endpoint.
Uncontrolled, prospective trials suggest that antiviral therapy may afford salutary effects in patients with HBV-related PAN.  The French Vasculitis Study Group performed an uncontrolled, open-label trial of glucocorticoids, plasma exchange, and vidarabine A therapy in 33 patients with PAN and HBV infection.33  After a 2-week course of prednisone, patients were treated with plasma exchange for several months and vidarabine for 3 weeks.  Circulating hepatitis B early antigen (HBeAg) and surface antigen (HBsAg) cleared after the first cycle of vidarabine in 12 (36%) and 5 (15%) patients, respectively.  A second cycle of vidarabine produced HBeAg clearance in 3 additional patients.  Of the patients treated by this protocol, 24 recovered completely from vasculitis.  Eight patients died:  three form uncontrolled vasculitis, one from fulminant hepatitis, and four from other causes.  Although these results are somewhat encouraging, vidarabine therapy is no longer used for the treatment of HBV infection because of its neuromuscular toxic effects and its failure to sustain long-term benefits.
More recently, IFN-alpha has been shown to be effective for the treatment of chronic hepatitis B.  The results of a meta-analysis involving more than 800 patients with chronic hepatitis B show that IFN-alpha treatment was associated with disappearance of serum HBeAg and serum HBV DNA in 33% and 37% of patients, respectively.46  Among the placebo recipients, HBeAg seroconversion occurred in 12%, and serum HBV DNA disappeared in 17%.  The immunomodulatory effects of IFN-alpha are complex and may theoretically hasten the progression of vasculitis.  The results from a small prospective trial involving six patients with HBV-related PAN suggested that plasma exchange combined with 3 MU of IFN-alpha administered three times a week can control the vasculitic manifestations of PAN.34  In this open-label study, HBeAg and HbsAg cleared in four and three patients, respectively.  The recommended IFN-alpha dose for the treatment of chronic hepatitis B is 5 MU/day for 16 weeks, higher than that used in the cited studies.
Lamivudine and famciclovir, oral nucleoside analogs, have been shown to inhibit HBV replication in vitro and in vivo.  The use of lamivudine in HBV-related PAN has not yet been reported.  In one patient with HBV-related PAN who was treated with a combination of IFN-alpha and famciclovir, HBV DNA and HBeAg were cleared from the circulation, and the patient showed clinical resolution of PAN.35
Long-Term Experience
Certain clinical features of PAN and CSS are associated with a worse prognosis.  The French Vasculitis Study Group prospectively studied 342 patients in five different clinical trials.  The investigators found that proteinuria of more than 1 g/day, a serum creatinine level exceeding 1.6 mg/dL, and gastrointestinal tract involvement were associated with a significantly higher mortality rate than patients without these risk factors.36  Fortin and colleagues37 also found in 45 patients with PAN that renal and cardiac involvement are independent predictors of increased mortality.  Older studies suggested that HBV infection in patients with PAN confers an increased risk of chronic liver failure.11,12,18
Relapses occur in approximately one third of patients with PAN who achieve initial disease control with a variety of treatment regimens.28,29,38 Relapses have been documented in 26% of patients with CSS.39  In most cases, these relapses have been preceded by a rise in the eosinophil count.39 Despite aggressive therapy, about 10% to 20% of patents with PAN and CSS die 5 to 10 years after disease onset from active disease or from a disease-or treatment-related complication.17,20,21,27-29 Most early deaths result from a perforated viscus, massive gastrointestinal hemorrhage, myocardial infarction, cerebrovascular accident, or renal failure caused by active disease.  Cardiac insufficiency, sudden cardiac death, and respiratory failure are frequent causes of death in patients with CSS.39 A minority die of infections during treatment.
PAN, MPA, and, in rare cases, CSS can lead to end-stage renal disease (ESRD).  Patients who become dialysis-dependent at the outset of treatment may nonetheless recover renal function and exhibit favorable renal outcomes.40  Proteinuria in the range of 3 to 5 g/day has been associated with the development of chronic renal failure.  ESRD developed in 61 (20%) of 308 patients with ANCA-associated systemic vasculitis (WG, MPA, isolated necrotizing glomerulonephritis).  These patients later received either chronic dialysis or a kidney transplant and had survival rates of 74% at 1 year and 40% at 5 years,41 which were similar to the survival rates of the general dialysis population in a U.S. national registry.  In this group of vasculitis-induced ESRD patients, vasculitic relapses occurred in 2 of the 22 patients with a renal transplant who were receiving maintenance immunosuppressive therapy and in 7 of 37 patients undergoing dialysis.  The flares in these instances generally responded to increaded doses of glucocorticoids or reinstituion of cyclophosphamide.  Other small series show that patients with ANCA-associated MPA can undergo successful renal transplantation with a low rate of subsequent vasculitic relapse.42, 43
The treatment of PAN and MPA was reviewed by Guillevin and Lhote.44  Patients with active PAN, MPA, or CSS are treated with high doses of glucocorticoids.  Some judgment can be exercised in deciding whether to initially combine these drugs with a cytotoxic agent.  In my views, serious visceral organ system disease—such as mononeuritis multiplex, central nervous system deficits, congestive heart failure, renal involvement, or evidence of mesenteric ischemia—justifies the addition of oral cyclophosphamide to the initial treatment regimen.  Patients with PAN who lack serious visceral organ involvement can be managed initially with glucocorticoids alone and, if immediate improvement does not occur, subsequently treated with oral cyclophosphamide.
Certain factors may modify this treatment approach.  For example, a concurrent diabetic state may support the initial use of cytotoxic therapy to allow more rapid reduction in the glucocorticoid dose and better control of hyperglycemia.  Conversely, other considerations, such as diagnostic uncertainty, compliance issues, or a patient’s concern about drug side effects, may militate against use of a cytotoxic agent.
The glucocorticoids most commonly prescribed are prednisone and methylprednisolone (orally or intravenously) in starting doeses of 1 to 2 mg/kg/day.  A protocol for initiating glucocorticoid therapy without concomitant cytotoxic agents in systemic vasculitis is shown in Table 27-2.  Treatment is begun in rapidly progressing disease with divided daily dodese of glucocorticoids, which are more immunosuppressive and potentially toxic than equivalent doses of the same preparation given once a day.  A 70-kg person, for example, would be treated initially with prednisone or methylprednisolone in dosages of 15 to 30 mg four times/day.  Administration of glucocorticoids is then converted from divided daily doses to a single morning dose over the next 7 to 10 days.  With clinical improvement, the prednisone dose can be tapered over the first 2 weeks of therapy to a single morning dosed of 60 to 80 mg and over the next 3 to 4 months to 20 mg. The daily predisone dose may then be converted to an alternate-day schedule, an approach I favor, or reduced further in small decrements (e.g., 2.5 mg) until reaching the lowest dose that effectively controls the disease.  Concomitant cytotoxic therapy usually allows both a more rapid reduction in prednisone does, starting about 4 weeks after initiation of therapy, and an earlier conversion of the prednisone regimen from a daily dose to an alternate-day schedule (e.g., at 3 to 4 months).
The initial dosage of cyclophosphamide for treatment of systemic vasculitis is 2 mg/kg/day (usually given orally);  this dose is later adjusted, depending on the peripheral leukocyte count (to be described).  More fulminant vasculitis may warrant initiation of cyclophosphamide at a dose of 4 mg/kg/day for 3 days, followed by a dosage reduction within 7 days to 2 mg/kg/day.  Azathioprine (2 mg/kg/day) is far less effective than cyclophosphamide in inducing remission but may serve as an alternative in this situation.21  
Disease refractory to cyclophosphamide therapy may be managed by increasing the daily dose of cyclophosphamide in 25-mg increments until the peripheral leukocyte and neutrophil counts approach 3000/mm3 and 1500/mm3, respectively (to be discussed).  Sometimes higher doses of cyclophosphamide are tolerated when the prednisone dose is increased, preferably on the alternate-day schedule.  A second option is alternative or experimental therapy (to be described).
The tolerance of the myeloid compartment to cyclophosphamide usually wanes with prolonged treatment because of diminishing bone marrow reserves.  The cyclophosphamide dose is reduced only if the peripheral leukocyte count drops below 3000/3 x 109 /L or the absolute neutrophil count falls below 1500/1.5 x 109/L.  Azathioprine may sustain remission if unacceptable side effects result from oral cyclophosphamide therapy.38,45  The patient should receive oral cyclophosphamide for 12 months after a complete remission, followed by a gradual withdrawal of the drug over 2 to 3 months.
Cutaneous PAN generally necessitates less immunosuppressive because of its lack of systemic manifestations and favorable long-term prognosis.  Local measures and high-po tency topical or intralesional glucocorticoids may be tried as the first approach.  Salicylates, dipyridamole, sulfapyridine, pentoxifylline, and dapsone are only variably effective in the treatment of cutaneous PAN,46  but they may be useful in individual cases.  Severe, rapidly progressive skin disease can respond to prednisone in doses of up to 1 mg/kg/day.
Some patients with chronic skin lesions may not improve with glucocorticoids or may suffer excessive complications from the prolonged use of these drugs.  In this situation, low-dose weekly oral methotrexate (7.5 to 20 mg/week) may promote healing of skin lesions and act as a steroid-sparing agent.47  Although oral cyclophosphamide is effective in the treatment of systemic vasculitis, it less consistently induces remissions of cutaneous PAN.  Moreover, combining oral cyclophosphamide with daily doses of glucocorticoids can predispose the patient to bacterial infections of nonhealing skinlesions.  Thus cytotoxic agents are used with caution in cutaneous PAN.  Because the major clinical manifestions of cutaneous PAN are painful skin ulcers, local therapy aimed at protecting the wound, decreasing superficial infection, and promoting healing is often the most productive avenue of therapy.
The treatment approach must be modified in cases of HBV-related PAN because excessive immunouppression may worsen outcomes.  The vasculitic manifestation may be brought under initial control with a short course of prednisone.  According to published studies, successful outcomes have been achieved by starting prednisone at a dose of 1 mg/kg/day for 1 week and stopping it by the end of the second week after a rapid taper.  The treatment of HBV-related PAN should include an antiviral agent.  IFN-alpha has emerged as the drug of choice for HBV infection.  The recommended dose of IFN-alpha is 5 MU every day for 16 weeks.  However, IFN-alpha has only been used in patients with HBV-related PAN at doses of 3 to 5 MU three times a week, the usual doses for treating hepatitis C.  The treatment duration depends on the clearance of serum HBeAg.  IFN-alpha injections are discontinued after loss of serum HBeAg, whereas persistent HBe antigenemia calls for continuation of the IFN-alpha treatments for up to 1 year.  There does not appear to be a raionale for using less than the recommended dosage unless the patient has difficulty tolerating that dosage.
      IFN-alpha should be considered as first-line therapy in patients with low HBV loads (e.g., low HBV DNA levels), in whom the cure rate has been the highest.  IFN-alpha treatment offers a low probalility of cure in patients with high pretreatment HBV DNA levels.  In such cases, lamivudine, 100 mg/day, would be a reasonable alternative despite the lack of published studies in which it is administered to patients with HBV-related PAN.  Lamivudine has been well tolerated by most patients and significantly reduces serum HBC DNA levels in almost all cases of chronic hepatitis B infection.  However, treatment with lamivudine has little effect on the cure rate in chronic hepatitis B regardless of viral load.
Another treatment option for HBV-related PAN is plasma exchange.  The therapeutic value of plasma exchange has not been shown in controlled trials.  Plasma exchange may serve to control the acute and subacute manifestations of vasculitis, and the antiviral agent acts to eliminate the triggering antigen (e.g., HBV).  Plasma exchange may be initiated after tapering of the prednisone therapy, beginning three times weekly for the first 6 weeks, followed by twice weekly for 3 weeks and then once or twice every week according to the clinical response.  One plasma volume (30 mL/kg x [100 – hematocrit]) is removed per session and replaced with lactated Ringer’s solution and albumin.  Fresh-frozen plasma is used as needed to replenish coagulation factors.
TABLE 27-2
A Protocol for Initiating Glucocorticoid Therapy Without
Concomitant Cytotoxic Agents in Systemic Necrotizing Vasculitis
Glucocorticoid Dose & Time From Onset
Preparation Route Frequency Of Therapy
Methylprednisolone IV 20 mg q6h Day 1
Methylprednisolone IV 40 mg q12h Day 5
Prednisone Oral 60 mg/d Day 10
Prednisone Oral 60 mg/d Week 4
Prednisone Oral 40 mg/d Week 8
Prednisone Oral 20 mg/d Week 12
Prednisone Oral 40 mg, alt w/30 mg/d Week 17
Prednisone Oral 40 mg, alt w/20 mg/d Week 18
Prednisone Oral 40 mg, alt w/10 mg/d Week 19
Prednisone Oral 40 mg QOD Week 20
   alt w/= Alternating with; IV=intravenous
Clinical Pharmacology and Drug Interactions        
Glucocorticoids produce a myriad of biologic effects.  After moving across the cell membrane, they complex with high-affinity receptors and enter the nucleus, where they change the transcription of specific genes.  The products of these genes regulate diverse intracellular and extracellular events associated with cellular activation and inflammation.  Glucocorticoids interfere with endothelial cell function, inhibit recruitment of circulating leukocytes to sites of tissue inflammatio, and block proinflammatory mediator production.  Pharmacologic doses of glucocorticoids suppress the hypothalamic-pituitary-adrenal axis, reducing endogenous cortisol production.
Glucocorticoid preparations differ in relative potency, sodium-retaining properties, and biologic half-life.  Those most commonly used to treat vasculitis are the short-acting preparations: prednisone and methylprednisolone.  Because of their longer biologic half-life, the longer-acting preparations, such as dexamethasone, exhibit a much greater “steroid effect” than do equivalent doses of dexamethasone, for example, suppresses cortisol production for more than 48 hours, and thus dexamethsone is not suitable for long-term daily administration.
Because glucocorticoids are inactivated in the liver, drugs that induce hepatic microsonal enzyme activity, such as phenytoin, barbiturates, and rifampin, may accelerate the elimination of glucocorticoids.48   Glucocorticoids may also speed the metabolism of salicylates in the liver, lowering the serum salicylate levels.48  Pharmacologic doses of glucocorticoids can accentuate the need for antihypertensive medications or increase the insulin requirements of patients with diabetes mellitus.
Cyclophosphamide is a potent immunosuppressive drug.  It inhibits cellular activity by alkylating DNA, which crosslinks the DNA coil and introduces coding errors during DNA synthesis and transcription.  Thus rapidly dividing cells are its principal target.  In does used to treat vasculitis, cyclophosphamide inhibits lymphocyte proliferation more strongly than neutrophil proliferation, resulting in a greater degree of lymphopenia than neutropenia.  Cyclophosphamide also down-regulates both B cell and T cell function.
Hepatic microsonal enzymes are involved in the activation of cyclophosphamide.  Drugs that induce these enzymes may alter the conversion of cyclophosphamide from its inactive to its active form.  Of more importance is that the elimination and detoxification of cyclophosphamide may be prolonged in patients with liver or renal dysfunction.
Drug Side Effects and Recommended Monitoring  
Glucocorticoid complications are more pronounced with longer acting preparations, more frequent adminstration, and prolonged use.  Physicians should be alert for the occurrence of these toxic effects.  The initiation of high-dose glucocorticoids may produce sudden and dramatic hyperglycemia in susceptible persons.  Patients should be warned about polyuria and polydipsia.  An intervening serious illness or a surgical procedure may warrant temporary administration of “stress doses”  of glucocorticoids because of hypothalamic-pituitary-adrenal suppression.  Unless contraindicated, 400 IU of vitamin D and 1000 to 1500 mg of elemental calcium can be given daily to retard the development of glucocorticoid-induced osteoporosis.  Alendronate sodium or calcitonin may also be indicated for prevention of glucocorticoid-induced osteoporosis.
Cyclophosphamide therapy is associated with substantial toxicity.  The most common side effects are dose related and include gastrointestinal intolerance, hair loss, and myelosuppression.49  Its myelosuppressive actions can be a major factor in the development of serious infections.  Patients receiving cyclophosphamide have frequent infections from common bacterial pathogens, opportunistic organisms, and herpes zoster; these infections occur most often in patients who are receiving concurrent daily dosages of prednisone.50, 51  Keeping the prednisone dosage on an alternate-day schedule appreciably lowers the risk of infection.
The cyclophosphamide dosage must be adjusted on an idividual basis to avoid excessive myelosuppression.  With renal impairment (e.g., serum creatinine level > 2 mg/dL), the initial dosage can be reduced by 25% to 50%, although the level of renal function does not reliably correlate with the maximum tolerated dose.  The cyclophosphamide dose is held constant for the first 10 to 14 days of therapy, with subsequent dosage adjustments to keep the total leukocyte count above 3000/3 x 109 /L and the neutrophil count above 1500/1.5 x 109 /L.  The peripheral leukocyte count begins to drop 8 to 14 days after cyclophosphamide is initiated or its dosage is increaded.  Thus measured counts in peripheral blood reflect suppression of the marrow, not from the current dose, but rather form the dose taken 1 to 2 weeks earlier.  Waiting until the leukocyte counts drop below safe limits almost invariably results in prolonged myelosuppression.  Such unwanted toxicity can be avoided by monitoring the rate of decline in the peripheral leukocyte count, anticipating its nadir, and making the appropriate dosage adjustment before the full myelosuppressive effects of the current dose are realized.  Recovery from myelosuppression takes 18 to 25 days.
Glucocorticoids influence the numbers of circulating neutrophils by stimulating their release from the bone marrow and inhibiting  their adhesion to vascular endothelium.  Treatment with high doses of glucocorticoids often increases the peripheral leukocyte count to the range of 10,000 to 20,000/10-20 x 109 /L.  About 2 weeks into the treatment of systemic vasculitis, the usual practice of consolidating the glucocorticoid dose to once a day, combined with the delayed myelosuppressive effects of cyclophosphamide, can result in a rapid fall in the leukocyte count that must be closely monitored to avoid excessive immunosuppression.  Cyclophosphamide overdosage can be averted by checking the peripheral leukocyte count at least twice weekly for the first 3 weeks of therapy (or possibly longer until the counts level off in a safe range) and making prompt dosage adjustments.  Subsequently, the peripheral leukocyte counts can be monitored on a weekly basis and every 2  weeks after 2 months of constant dosing.
The teratogenicity of cyclophosphamide mandates that effective birth control practices be maintained during the treatment period.  Cyclophosphamide intereferes with the development of oocytes and spermatocytes and can promote gonadal fibrosis and clinical sterility.  These gonadal effects frequently produce oligomenorrhea and amenorrhea in women and decreased libido, impotence, and azoospermia in men.  The risk of permanent sterility appears to be dose related and higher in women over 30 years of age.52 Concomitant use of oral contraceptives by women during cyclophosphamide therapy may reduce the incidence of sterility.53  Before initiation of treatment with cyclophosphamide, male patients should be advised about banking sperm to aid future conception.
Acrolein, a toxic metabolite of cyclophosphamide, can irritate the bladder and promote hemorrhagic cystitis, bladder fibrosis, and transitional and squamous cell carncinoma.49  Two preventive measures may lessen the frequency of bladder complications.  One widely accepted approach is maintenace of a liberal fluid intake of several liters daily.  Mesna (2-mercaptoethanesulfonate) has also been advocated for prophylaxis54 and is available for intravenous use.  The parenteral solution can be taken orally because it is 29% to 81% bioavailable by this route and stable for up 24 hours in a variety of beverages.55 
The urine should be examined every month during cyclophosphamide therapy to monitor for signs of bladder toxicity.  Unexplained hematuria, unless it clears rapidly, mandates an immediate diagnostic evaluation, including urology consultation and cystoscopy.  The occurrence of severe cystitis usually necessitates discontinuation of oral cyclophosphamide.  On the other hand, mild cystitis that resolves quickly can be managed by reemphasizing the need for good hydration and resuming the cyclophosphamide at a 25- to 50-mg lower dosage.  Hematuria also can signal a neoplastic change in the bladder mucosa.  To screen for bladder cancer, a urine cytologic examination is performed yearly in patients whose total cyclophosphamide dose exceeds 10 g.  Neoplastic changes have been detected in the bladder mucosa up to 11 years after the completion of cyclophosphamide therapy.54  Thus, after cyclophosphamide therapy, patients must have annual surveillance for the potential development of bladder neoplasms.  An overall scheme for monitoring oral cyclophosphamide therapy is outlined in Table 27-3.
A major concern of long-term cyclophosphamide therapy is delayed appearance of malignant disease, mostly leukemias and lymphomas.56  Because the risk of these secondary malignant conditions is greater with higher total doses of cyclophosphamide, the need for this drug should be reexamined in patients who have received it for more than 2 years.
TABLE 27-3
Recommended Monitoring of Patients Receiving Oral Cyclophosphamide for 
the Treatment of Systemic Necrotizing Vasculitis
 Weeks 1-3 CBC twice a week Carefully note the rate of decline in 
  Periheral leukocyte count;  adjust 
 Cyclophosphamide does to keep 
  Leukocyte count >3000/3 x 109/L and 
  Neutrophil count >1500/1.5 x 109 /L
Weeks 3-8 CBC every week As for weeks 1-3
Urinalysis every month Hematuria may indicate cystitis
Week 9 until end of CBC every 2 weeks As above
   therapy   Urinalyisis every month As above
Unirne cytology study Screen for bladder cancer
  Every year
After therapy Urine cytology study every Maintain surveillance for bladder cancer
  Year   indefinitely
*CBC = complete blood cell count.  The period after initiation of cyclophosphamide therapy.
Investigational Drugs and Future Directions
Alternative approaches have been used frequently in patients with systemic vasculitis.  Intravenous pulse glucocorticoid therapy (e.g., methylprednisolone, 0.5 to 1 g/day for 3 days) has shown efficacy in the treatment of PAN and CSS57,58  and may be useful as a therapeutic alternative in certain clinical situations (e.g., for patients receiving anticoagulation therapy) or in refractory disease. Pulse cyclophosphamide therapy (500 to 750 mg/m2  given IV every month for 6 months) may produce therapeutic benefits in PAN and CSS.30-32,59,60  Pulse cyclophosphamide therapy can be administered as a single intravenous bolus, as routinely done in the management of lupus nephritis, or as multiple oral doses over 3-day period.38  Whether pulse cyclophosphamide can substitute for daily oral cyclophosphamide in the treatment of systemic vasculitis remains controversial.
Cyclosporine is a powerful immunoregulatory agent that can effectively suppress active systemic vasculitis, although it may be frequently toxic at the dosages necessary to control disease.61  Treatment with cyclosporine has been inititated at dosages of 5 mg/kg/day and, depending on the clinical response and tolerability, gradually reduced to doses of 1 to 3 mg/kg/day for long-term maintenance.
Treatment with a combination of two monoclonal antibodies transiently reduced the number of circulating T cells and improved clinical disease manifestations in four patients with refractory systemic vasculitis. 62, 63  Innovative therapies such as this will undoubtedly be the focus of more research in the future.  The approval by the US Food and Drug Administration of etanercept, a p75-soluble tumor necrosis receptor fusion protein (TNFR:Fc), will almost certainly stimulate future studies of tumor necrosis factor-alpha blockade therapy for PAN, MPA, and CSS.
WG is a necrotizing vasculitis with a predilection for the upper and lower respiratory tracts and kidneys.  Otitis media, mucosal ulceration, nasal perforation and deformity, and sinusitis dominate the manifestations of upper airways involvement.  The most common sysmptoms of lower airway disease are cough and hemoptysis.  In addition, symptoms of subglottic stenosis occurred in 43 (23%) of 189 patients with WG followed at the National Institutes of Health (NIH).64   Chest radiographs of patients with symptoms as well as of those who have no symptoms may show bilateral nodular infiltrates and cavities.  Other features of WG include ocular inflammation, retroorbital masses, arthritis, multiple mononeuropathies, cerebral arteritis, and vasculitic skin lesions.
WG has a variable course, depending on the sites of clinical involvement.  The presence of both respiratory and renal disease has been termed generalized WG, which, without aggressive immunosuppressive therapy, augurs a poor prognosis.  A less ominous course is taken by patients with limited WG, a subgroup with respiratory disease and no clinical evidence of renal or other serious organ system disease.65  
The erythrocyte sedimentation rate (ESR) is positively correlated in WG with clinical disease activity.  However, caution must be exercised in interpreting this test result because concomitant infection, such as sinusitis, may also elevate the ESR.  ANCA with a granular cytoplasmic staining pattern (c-ANCA) have diagnostic significance in WG.  They occur in sera from 88% to 96% of patients with active, generalized WG.66, 67 The frequency of a positive c-ANCA test result drops to 67% among patients with locally active disease67 and to 41% to 43% among patients with inactive disease.66 c-ANCA occur infrequently in association with MPA, idiopathic rapidly progressive glomerulonepritis, autoimmune hepatitis, uveitis, amebiasis, fungal infection, and human immunodeficiency virus (HIV) infection and must not be used as the only criterion for diagnosis.66,68  Proteinase 3 (PR3), a major autoantigen in WG, is constituent of azurophilic granules in neutrophilis.  The presence of serum antibody to PR3 is highly specific for the diagnosis of WG.  The c-ANCA titer and levels of serum PR3 anitbodies have been used to monitor disease acitivity in WG.  Some studies have shown that a rise in the c-ANCA titer or level or anti-PR3 antibodies over time may correlate with increased clinical activity or may be the harbinger of a relapse.69  However, the evidence to date does not support the use of this serologic marker as a sole basis for initiating or increasing immunosuppressive therapy.69 
Depending on the biopsy site, areas of disease involvement in WG may exhibit histologic evidence of necrotizing granulomatous vasculitis, necrotizing granuloma without vasculitis, or acute and chronic inflammation.  The diagnosis is most often confirmed by the pathologic findings from an open-lung biopsy, which are also valuable for ruling out infection and other conditions that may resemble WG such as lymphomatoid granulomatosis or lymphoma.  Other biopsy sites may be equally informative, although bipsy specimens from upper airways are notorious for yielding nondiagnostic material.  Brochoscopy with tracheobronchial mucosal biopsy may be useful in evaluating lower airway symptoms such as cough, dyspnea, wheezing, or hoarseness.  Of 21 transbronchial biopsies from 17 patients with WG and airway symptoms, results of 3 specimens from 3 patients were diagnostic for WG and results of 9 specimens from 7 patients were compatible with WG.70 Renal biopsies show histologic evidence of a pauci-immune focal, segmental, necrotizing glomerulonephritis.
Nearly all patients with WG and upper airway involvement suffer from recalcitrant sinus infections.  Patients with WG are predisposed to chronic sinus infections because of the damage to the nasal mucosa from acute and chronic inflammation.  The organism to infect the paranasal sinuses most frequently is Staphylococcus aureus.  The nasal carriage of S. aureus is considered to be a risk factor for the development of S. aureus infections, and chronic nasal carriage of S. aureus in patients with WG may define a subgroup who are at high risk for disease relapse.71 The role of S. aureus and other microbial pathogens in triggering disease relapse is unclear.
Natural History of Disease
The initial clinical and histopathologic descriptions of patients with WG portray a rapidly fatal course without therapy.  Such patients survive a mean of 5 months and die of uremia, respiratory failure, or disseminated vasculitis.  Some patients with limited WG survive longer and die later of respiratory complications or progression to renal failure.
Uncontrolled Treatment Studies and Long-Term Experience
Although treatment with glucocorticoids extends the mean survival of WG patients to 12 months, vastly improved outcomes followed the adoption of oral cyclophosphamide as standard therapy. Researchers from the NIH summarized their experience in treated 158 patients with WG who had been monitored for 1229 patient-years. Nearly all the patients group had received initial treatment with 1 mg/kg/day of prednisone and 2mg/kg day of oral cyclophosphamide, with subsequent reduction of the prednisone dose to an alternate-day schedule and maintenance of daily oral cyclophosphamide for 12 months after a complete remission. With this therapy, 91% of the patients showed clinical improvement, and 75% attained a complete remission. About half of the group had at least one relapse and nearly the same number sustained a remission that lasted more than 5 years.
Disease-and treatment-related morbidity were common in this patient population and included chronic sinus dysfunction (47%), chronic renal insufficiency (42%), hearing loss (35%), nasal deformities (28%), pulmonary insufficiency (17%), tracheal stenosis (13%), and visual loss (8%). Complications attributed to cyclophosphamide were cystitis (43%), hair loss (17%), bladder cancer (2.8%), and myelodysplasia (2%). Overall, the rate of bladder cancer was 33 times higher than in the general population. Non-Hodgkin’s lymphoma developed in two patients; this rate represented an 11-fold increase in overall risk. Glucocorticoid therapy was associated with transient cushingoid features (in most patients), cataracts (21%), fractures (11%), diabetes mellitus (8%), and avascular necrosis (3%). Slightly more than half of the women of childbearing age could not become pregnant of had ovarian failure after more than 1 year of cyclophosphamide therapy. Nearly half of the patients required at lease one hospitalization for treatment of a serious infection, usually bacterial. Eight fungal, six Pneumocystis carinii, and two mycobacterial infections, as well as 34 episodes of herpes zoster, occurred during follow-up. 
Of the patients in this study, 20% died. Death was caused by active WG or a disease-or treatment-related complication in 13% of the patients; other deaths resulted from renal disease (3%), pulmonary disease (3%), pulmonary disease (1%), infection (3%), and cancer (2.5%).
Progression in ESRD occurred in 11% of the NIH patients. ESRD developed in 27 (22%) of 123 patients in a selected group with renal involvement secondary to WG or MPA. Extrarenal relapses of disease occur in a minority of patients during dialysis. The relapses have been readily treated with high doses of glucocorticoids but, in some cases, have necessitated addition of cyclophoshamide. The experience with renal transplantation in WG has been generally favorable, with only rare recurrences of disease in the transplanted organ and maintenance of graft function at a rate comparable with that of the total transplant population. 
WG may manifest during pregnancy and may necessitate aggressive immunosuppressive therapy. According to case reports, excellent outcomes have been achieved during pregnancy despite the need to treat active disease with glucocorticoids and cyclophosphamide. The mothers have usually improved clinically with an aggressive treatment approach, and their infants have had relatively few complications. 
Prospective Clinical Trials
The high rate of treatment-related complications from daily oral cyclophosphamide therapy has stimulated a search for less toxic immunosuppressive regimens. In an open-level, prospective trial, 14 patients with WG were treated with IV monthly pulses of cyclophosphamide, 1 g/m. Of the 14 patients, 12 had previously received glucocorticoids and oral cyclophosphamide. In this study, pulse cyclophosphamide therapy was associated with a rate or relapse higher than that in historical controls who had been treated with daily oral cyclophosphamide. These results are similar to those of other prospective and retrospective studies.
The relative benefits of pulse cyclophosphamide over daily oral cyclophosphamide were evaluated in a randomized, controlled trial involving 50 patients with newly diagnoses WG. After 6 months, 24 (90%) of the patients receiving oral cyclophosphamide therapy achieved remission, as opposed to 18 (78%) patients receiving oral cyclophosphamide; this difference was not statistically significant. However, the relapse rate was significantly higher among patients receiving pulse cyclophosphamide than among those receiving oral cyclophosphamide (52% vs. 18%; P<.05). Infectious complications were more frequent in those receiving daily oral cyclophosphamide than in those receiving pulse cyclophosphamide. In particular, P. carinii pneumonia was diagnosed in 30% of patients receiving oral cyclophosphamide but in only 11% of pulse cyclophosphamide-treated patients, which prompted the investigators to institute routine prophylaxis with trimethoprim-sulfamethoxazole (T/S) during the trial. Deaths were more common in this prospective study than in previous uncontrolled studies. After more than 2 years of follow-up, the mortality rates were 33% (9 deaths) and 44% (10 deaths) in the pulse and oral cyclophosphamide arms, respectively. These rates are much higher than the 20% mortality rate noted in the NIH series. Some of the excessive deaths in the clinical trail resulted from serious infections, which caused 3 (33%) of the deaths in the pulse cyclophosphamide recipients and 6 (60%) of the deaths in the oral cyclophosphamide recipients. 
A few points about this trial deserve further comment. Although treatment with pulse and oral cyclophosphamide in this trial produced similar survival rates, the two therapies should not be judged equivalent, because of the high probability of type II error. The unusually high rate of serious infectious may have come from the use of relatively high dosages of glucocorticoids. According to the protocol, the patients were treated initially with methylprednisoline, 15 mg / kg / day for 3 days, then 1 mg / kg / day on day 4 for 6 weeks, followed by a 2.5 mg dose reduction every 10 days, which translates to a dose of 15 mg / day at 6 months after initiation of therapy. In comparison the NIH group reduced the prednisone dosage in their patients to alternate-day therapy (e.g., 60 mg every other day) within a mean of 3.2 months. These data provide further support for the concept that an expedient reduction of the prednisone dose to alternate-day therapy significantly lowers the incidence of infectious complications. Further studies are required before pulse cyclophosphamide therapy can be accepted as an initial standard treatment for WG.
More promising results have been obtained with methotrexate in a subset of WG patients without immediately life-threatening disease. In an open-label prospective study, oral weekly methotrexate, combined with standard glucocorticoid therapy, produced remission in 30 of 42 patients with WG fitting this clinical profile. Of the 42 patients, 26 had had prior therapy with glucocorticoids, T/S, or daily oral cyclophosphamide. The dosage of methotrexate was begun at 0.3 mg / kg / week (not exceeding 15 mg / week) and increased to 20 to 25 mg / week as tolerated. Although the majority sustained remission, 11 (36%) of the 30 patients achieving remission with methotrexate therapy subsequently suffered relapse. Seven patients developed serious toxic effects that necessitated discontinuation of the methotrexate therapy subsequently suffered relapse. Seven patients developed serious toxic effects that necessitated discontinuation of the methotrexate therapy, including four cases of P. carinii pneumonia and three cases of methotrexate pneumonitis. These results suggest that weekly methotrexate therapy may play a role in the treatment of patients with non-life threatening manifestations of WG (e.g. limited WG). 
Considerable interest has focused on the role of T/S in the treatment of WG. Because infection has long been suspected to trigger WG, any beneficial effects from T/S could stem from its antimicrobial properties. De Groot and coworkers compared the clinical efficacy of methotrexate along, T/S alone, methotrexate plus prednisone, and T/S plus prednisone in 65 patients with WG who had achieved remission with conventional therapy. Approximately 85% to 90% of the patients maintained remission in the two groups receiving methotrexate, whereas nearly 50% of the patients taking T/S suffered relapse. Although methotrexate was superior to T/S in this study for maintaining remission, the strength of the evidence is relatively weak because of the nonrandomized study design, the inclusion of some patients in the trial who had not attained a complete remission, and the unknown effects of previous induction therapy.
The therapeutic value of T/S was further investigated in a 24-month, prospective, randomized, placebo-controlled trial involving 81 patients with WG whose disease had been brought into complete remission through standard therapy. T/S (160 mg of trimethoprim, 800 mg of sulfamethoxazole) given twice daily significantly reduced the rate of relapse, in comparison with placebo. Eight patients (20%) withdrew from the trial in the T/S group because of the side effects. The difference in relapse rate between the two treatment groups was significant only for recurrent disease in the nasal and upper airways and did not obviate the need for subsequent immunosuppressive therapy.
Most authorities stage the initial treatment of WG on the basis of the severity of disease. Glucocorticoids and another immunosuppressive drug are initially required for virtually all patients with active WG. Patients with severe WG should receive initial treatment with 1 mg / kg / day of prednisone (or its equivalent) and 2 mg / kg / day of oral cyclophosphamide. The prednisone dose is usually tapered to 20 mg / day by 2 months, to alternate-day therapy by 3 to 4 months, and stopped after 6 months. Oral cyclophosphamide is maintained for 1 year after the patients have achieved disease remission. Many authorities have begun to shorten the course of oral cyclophosphamide therapy to minimize treatment-related complications. Patients who achieve remission in 3 to 6 months with daily oral cyclophosphamide may be switched to weekly methotrexate therapy (described next). The methotrexate is continued until the disease has been in complete remission for 1 year, and the dose is then tapered by 2.5 mg / month. Patients with a complete remission and a serum creatinine level exceeding 2.0 mg / dL switch to azathioprine, 2 mg / kg / day, and are likewise treated for 1 year. The azathioprine may be subsequently tapered every 2 months in 50-mg dose decrements.
Methotrexate may substitute for oral cyclophosphamide in patients with limited WG. In such cases, the methotrexate dose is initiated at 0.25 mg / kg / week (maximum of 15 mg / week) and increased by 2.5 mg / week increments every week up to 0.35 mg / kg week (maximum dose, 25 mg / week) as tolerated. The methotrexate should be continued until the disease has been in complete remission for 1 year, and then the dosage should be tapered by 2.5 mg / month. Pulse cyclophosphamide, 500 to 750 mg / m. may be considered for initial or maintenance treatment of patients with WG who have contraindications to oral cyclophosphamide and methotrexate. Patients receiving glucocorticoids in combination with methotrexate or cyclophosphamide should be treated with T/S three times weekly to reduce the risk for P. carinii pneumonia. Alternatively, sulfa-allergic patients may be given dapsone, 100 mg / day, or inhaled pentamidine once a month. 
Patients receiving methotrexate therapy should take concomitant folic acid, 1 mg / day, or folinic acid, 5 mg / week, to reduce the risk of methotrexate toxicity. Methotrexate should be combined with therapeutic doses of T/S because together they may increase hematologic toxicity. Because methotrexate is excreted through the kidneys, its use should be avoided by patients with a serum creatinine level about 2 to 2.5 mg / dL. Patients with mild renal insufficiency who are treated with methotrexate should receive this agent in lower than usual doses (5-10 mg / wee) and be closely monitored for possible toxicity.
A disease flare necessitates intensification of immunosuppressive therapy. Patients with severe disease and a major flare who are receiving methotrexate or azathioprine may require reinduction with high dosages of prednisone and with oral cyclophosphamide therapy. Those with limited disease who develop severe flares are treated with increased doses of prednisone and may be switched from methotrexate to oral cyclophosphamide. Less severe flares in the context of limited disease may be managed by higher doses of methotrexate. Alternatively, mild-to-moderate flares of WG may be treated for 1 month with an increased dose of prednisone followed by resumption of a taper. 
Integral to the treatment of WG is aggressive treatment of sinus disease. Chronic sinus inflammation in WG impairs mucosal immunity and increases susceptibility to infection, usually from S. aureus. Frequent irrigation to relieve nasal crusting and blocked airways, surgical procedures to drain impacted sinuses, and prompt initiation of antibiotics at the earliest signs of infection are the cornerstones of managing chronic nasal and sinus disease. Worsening nasal or sinus disease that is unresponsive to local measures and antibiotic therapy may necessitate tissue biopsy for diagnosis. T/S in therapeutic doses may have a role in managing patients with smoldering sinus disease who are otherwise in remission.
Subglottic stenosis may mandate additional therapy if the patient is symptomatic. Intratracheal dilation and if the patient is symptomatic. Intratracheal dilation and instillation of corticosteroids has been shown to be a safe and effective option. Alternatively, successful repair of stenotic segments has been achieved with surgical reconstruction. 
Conjunctivitis, episcleritis, and anterior unveitis in WG are treated with tropical glucocorticoid preparations. In contrast, active scleritis and optic nerve vasculitis threaten vision and mandate immediate ophthalmologic evaluation to assess the need for systemic immunosuppressive therapy. A retroorbital mass, another worrisome ocular manifestation, can also jeopardize vision and may warrant emergency orbital decompression. 
Investigational Drugs and Future Directions
Several other therapies have been tried successfully in WG, including intravenous pulse methylprednisolone, cyclosporine, intravenous gamma globulin, etoposide, and monoclonal antibodies to CD52 or CD4 T cell antigens. Clinical trials are now under way with etanercept (soluble TNFR-Fc) in patients with WG because of the possible role of TNF-α in the pathogenesis of this disease.
Type II mixed cryoglobulinemia (MC) is characterized by recurrent purpura, arthralgias / arthritis, weakness, peripheral neuropathy, hepatomegaly, splenomegaly, Raynaud’s syndrome, Sjogren’s syndrome, and glomerulonephritis. The immunochemical composition of the cryoprecipitate defines the type of cryoglobulinemia. Type I cryoglobulins consist of a single monoclonal immunoglobulin and are found predominately in patients with malignancies of the immune system. Type II cryoglobulins are composed of a monoclonal immunoglobulin M (IgM) rheumatoid factor and polyclonal immunoglobulin G (IgG), whereas type III cryoglobulins contain polyclonal IgM rheumatoid factor and polyclonal IgG. Most of the type II cryoglobulinemias are essential, or primary. Secondary type II cryoglobulinemia has been associated with lymphoid malignancies, and secondary type III cryglobulinemia may occur with infections, autoimmune disease, and chronic liver disease. 
There is evidence indicating that hepatitis C virus (HCV) plays a major role in the pathogenesis of type II MC. Early series of trials with first-generation assays found a high prevalence of HBV infection in patients with type II MC, but later studies failed to confirm this association. More than 80% of cases of type II MC appear to be caused by HCV infection. Many patients with type II MC have no biochemical or clinical evidence of liver damage.
Histopathologic examination of skin lesions in type II MC reveals perivascular inflammation with varying degrees of leukocytoclasis. A study of lesional skin has shown by in situ hybridization that HCV as well as IgM and IgG localize in the epidermis, dermis, and the vessel wall. The glomercular pathologic process is characterized by focal or diffuse proliferative changes with intraglomerular thrombi IgG, IgM, and C3, the most prevalent immunoreactants, produce intense granular staining of peripheral capillary loops and massive staining of intraluminal thrombi. Electron miscroscopy shows electron-dense subendothelial deposits with a peculiar fibrillar or crystalloid structure. These morphologic patterns are also found in the serum cryoprecipitate and presumably represent high molecular weight aggregates of antigen-antibody complexes. Other involved organ systems show histopathologic evidence of small-vessel vasculitis.
Natural History of Disease and Long-Term Experience
The natural history of type II MC is poorly understood because virtually all patients with this condition have been treated with glucocorticoids and immunosuppressive therapy. Type II MC typically exhibits a chronic course and has a high probability of relapse. Episodes of purpura last 3 to 10 days and may recur sporadically over a period of years. Confluent purpura on the lower legs may evolve into deep skin ulcers, which are susceptible to infection and may progress to osteomyelitis.
Focal or diffuse proliferative glomerulonephritis develops in 8% to 54% of patients with type II MC. It may accompany the first episode of purpura or, more frequently, appear several years after the onset of skin and join disease. The renal lesion pursues a variable course. A partial or complete remission of renal disease occurs in nearly on third of affected patients, whereas in the others the disease typically progresses in an indolent manner or acute, reversible exacerbations occur. Earlier studies suggested that chronic renal failure develops commonly in type II MC; however, more recent studies have found that chronic renal failure develops in only about 10% of cases.
Most deaths in type II MC result from cardiovascular disease, cerebral hemorrhage, or infection. Cardiovascular deaths have occurred in patients with severe hypertension. Autopsies in two cases revealed evidence of coronary arteritis. Hypertension and cerebral arteritis are contributing factors in most cerebrovascular deaths. Infections of chronic, deep skin ulcers have led to death from sepsis.
Some investigators have postulated that HCV-related type II MC is a benign of low-grade B-cell lymphoproliferative disease. Bone marrow biopsy specimens from patients with type II MC have shown B-cell monoclonality as well as focal lymphoplasmacytoid infiltrates suggestive of non-Hodgkin’s lymphoma. In one study of type II MC, bone marrow biopsy revealed findings diagnostic or suggestive of non-Hodgkin’s lymphoma in 9 of 16 cases. Follow up of 200 patients with type II MC in another series disclosed 14 cases of non-Hodgkin’s lymphoma that had developed on average 5.6 years (SD, ± 3.8 years) after the original diagnosis. Four of these patients died of the malignancies. The rate of progression was even more striking in another group, in which non-Hogkin’s lymphoma developed in 12 (39%) of 31 patients with type II MC after long-term follow-up. The results from these studies confirm the presence of occult lymphoproliferative disorders in patients with HCV-related type II MC and show that some cases may progress to overt lymphoma. 
Uncontrolled Treatment Studies
The treatment of type II MC has evolved with our growing understanding of the pathogenesis of this disease. Initial therapies focused on the evidence that type II MC was an immune complex-mediated disease. Glucocorticoids and other immunosuppressive drug such as cyclophosphamide were the mainstays of therapy. Plasma exchange and cryofiltration have been used to rapidly remove cryoprecipitable immune complexes from the circulation, and they typically produce transient clinical responses. Disease manifestations usually rebound when plasma exchange is stopped without concurrent immunosuppressive therapy. Despite the diagnostic value of serum cryoglobulins in type II MC, uncontrolled observations show that improvement in disease activity after treatment does not always correlate with a reduction in the cryocrit. 
Outcomes are relatively poor in patients with type II MC treated with a combination of plasma exchange, glucocorticoids, and cyclophosphamide. In 1980, Popp and associates reported 18 deaths among 40 patients with type II MC despite aggressive combination therapy. Of the 20 patients with renal disease in this study, 14 died after an average of 7.4 years of follow-up.
Evidence that type II MC is probably caused by HCV infection has dramatically altered the management of this disorder. This first available anti-viral drug for drug for HCV infection was IFN-alpha. IFN-alpha treatment of chronic hepatitis B for 48 weeks produces a sustained virologic response in 15% to 20% of patients. Initial case reports suggested that IFN-alpha therapy for HCV-positive patients with type II MC decreased serum cryoglobulin levels and improved clinical disease activity. Longer-term studies of patients with type II MC treated with IFN-alpha showed sustained decreases in the level of viremia, reductions in cryocrit, and favorable clinical responses. 
Casato and colleagues prospectively analyzed the outcomes of 31 patients with HCV-positive type II MC who had received 3 MU of recombinant IFN-alpha daily for the first 3 months and then every other day for at least 9 additional months. Of the 31 patients, 17 (62%) achieved a complete remission after treatment with IFN-alpha. Only the presence of antibodies to C22 (which is an HCV structural protein antigen_ and low levels of HCV RNA before treatment were predictive of a complete response. Clinical responses were correlated with the disappearance of serum HCV RNA and of cryoglobulins and a decrease in the titer of anti C22 antibodies. Relapses also occurred in 5 of the 31 patients during IFN-alpha therapy. In 2 of the 3 patients monitored during follow-up, relapse was accompanied by the detection of serum antibodies to IFN-alpha. These 4 patients were then successfully treated with natural IFN-alpha, which suggests that treatment resistance may have resulted from the development of antibodies to recombinant IFN-alpha. 
Prospective Trials
The efficacy of IFN-alpha therapy has been examined in a randomized, crossover trial of 26 patients with type II MC. Patients were eligible for this trial only if they had hepatic or neurologic involvement and had no clinical evidence of renal disease. Most of the patients had been receiving low doses of prednisolone. HCV infection was documented in 25 of the 26 patients who entered this trial. IFN-alpha therapy was administered for 6 months (2MU / day for 1 month, then every other day for 5 months). Treatment with IFN-alpha was associated with statistically significant reductions in the purpura score, liver enzyme values and cryocrit. Upon stopping INF-alpha therapy, patients often experienced a rebound in disease manifestations.
A randomized, placebo-controlled, parallel trial was performed in 53 patients with HCV-related type II MC. In this study, patients were randomly assigned to receive treatment with placebo or IFN-alpha, 1.5 MU three times a week for 1 week and then 3 MU three times a week for the next 23 weeks. At the start of treatment, all 25 patients who received IFN-alpha and 25 of 26 patients who received placebo had detectable serum HCV RNA. The analysis was performed on only patients who completed the treatment course. IFN-alpha treatment significantly reduced the levels of viremia. Serum, HCV RNA levels became undetectable in 15 of the 25 patients who received IFN-alpha, in comparison with none of the 24 patients who received placebo. All the patients in whom HCV RNA disappeared from the serum showed clinical improvement and a markedly decreased level of serum cryoglobulins. Viremia recurred between 24 and 48 weeks after discontinuation of therapy in 13 of the 15 INF-alpha recipients and 1 year after discontinuation of therapy in the other 2 patients. In some cases, re-treatment with IFN-alpha induced a disease remission. Treatment-related side effects observed during the trial included flu-like symptoms after the first few injections (most patients), mild leukopenia and thrombocytopenia, loss of appetite, alopecia, depression, and insomnia. The dose of IFN-alpha had to be reduced in two cases because of thrombocytopenia. These data provide strong evidence that eradicating HCV infection may cure HCV-related type II MC.
The results of randomized, controlled trials involving patients with chronic hepatitis C indicated that treatment with the combinations of IFN-alpha and ribaririn produces a higher rate of cure than does treatment with IFN-alpha alone. Ribavirin, synthetic guanosine nuceleoside analog, has anti-viral activity in vitro against a wide range of RNA and DNA viruses. In these trials, IFN-α2b was administered at a dosage of 3 MU three times a week, and ribavirin was given orally at 1000 or 1200 mg / day. As initial treatment, sustained virologic and biochemical responses were observed in 35% to 43% of patients treated with the combination therapy for 24 to 48 weeks, in comparison with 18% to 19% of patients receiving IFN-alpha alone. The combination therapy also yielded higher rates of virologic and biochemical responses than did IFN-alpha alone in patients with chronic hepatitis B who suffered relapse after IFN-alpha treatment. The main side effect of ribavirin in these trials was hemolytic anemia (hemoglobin <10 g / dL), which was observed in 10% of patients receiving the combination therapy in clinical trials and may necessitate a decrease in the ribavirin dose.
There are no published studies of the combination therapy in patients with HCV-associated type II MC. However, five patients with HCV-positive type II MC have been treated in an open-label trial with 1000 to 1200 mg / day of ribavirin as monotherapy for 10 to 36 months. The virologic and clinical responses were transient in the five treated patients; thus, ribavirin therapy alone does not produce a sustained therapeutic effect in patients with HCV-positive type II MC. Other agents may hold promise. For example, treatment with a novel purine analog, 2-chlorodeoxyadenosine produced a remission in a patient with type II MC previously refractory to glucocorticoids, cyclophosphamide, plasma exchange, and IFN-alpha. 
The cornerstone of management is antiviral therapy, and the goal is to eradicate HCV from the body. Patients with moderate or severe clinical manifestations of small-vessel vasculitis may be treated initially with plasma exchange and low dosages of glucocorticoids to bring the disease quickly under control. However, the dosage of glucocorticoids should be tapered as rapidly as possible after the introduction of antiviral therapy. Local measures such as compressive stockings or Unna boots may be indicated for the management of purpura and skin ulcers. Immunosuppressive drugs such as cyclophosphamide should be strictly avoided in patients with chronic HCV infection. 
Studies have shown that most patients with HCV related type II MC benefit clinically from IFN-alpha therapy. The treatment protocols typically call for 3 MU of IFN-alpha (Intron A), given three times a week for 6 to 12 months. Although clinical responses to IFN-alpha monotherapy have been favorable, the cure rate appears to be only 15 % to 20 % with this approach. Higher cure rates may be possible through the combination of IFN-alpha and ribavirin for 24 to 48 weeks. However, no published experience with the combination therapy in patients with HCV-related type II MC is available. In view of the goal of antiviral therapy, the combination therapy would probably produce superior long-term outcomes in comparison with IFN-alpha alone and should be considered in patients with HCV-related type II MC, especially those with chronic hepatitis.
IFN-alpha therapy has been associated with severe depression and suicidal behavior and should be used with extreme caution in patients with a history of previous psychiatric illness. Because ribavirin often causes significant hemolytic anemia, all patients receiving the combination therapy should have a complete blood count at weeks, 1,2 and 4 after starting treatment and monthly thereafter. A liver biopsy should be considered for any patient with HCV infection and biochemical evidence of hepatic injury to monitor the response to antiviral therapy.
HSP is a multi-system disease that affects primarily the skin, joints, gastrointestinal tract, and kidneys. The ACR has developed criteria for the classification of HSP. This illness is most common in children. The key clinical feature is a purpuric rash that involves the lower extremities and buttocks. The majority of patients experience join pain and swelling. Abdominal pain commonly occurs in HSP and may be complicated by gastrointestinal hemorrhage or, in rare cases, intus-susception. Some degree of renal involvement, ranging from microscopic hematuria to acute nephritis and nephrotic syndrome, develops in approximately 40 % patients.
In addition to the urinary findings, laboratory studies may reveal mild leukocytosis and an elevated ESR. A characteristic clinical picture with a normal platelet count is usually sufficient for a clinical diagnosis. Biopsy of the purpuric skin lesions reveals a leukocyctoclastic vasculitis and prominent immunoglobulin A deposition in the vessel walls, which confirm its vasculitic nature.
Natural History of Disease
The acute illness usually lasts about 4 weeks, but more than one third of the cases persist longer. Recurrent episodes of HSP occur in 12 % to 40 % of children and are reminiscent of the initial clinical presentation, although milder. Age greater than 5 years at presentation has been associated with increased severity of renal disease. 
Kidney involvement accounts solely for the long-term morbidity and mortality in HSP. Most studies of unselected HSP cases suggest that progression to chronic renal failure occurs in fewer than 5 %. Kobayashi and coworkers retrospectively classified 203 Japanese children with HSP according to renal findings at presentation. They found no abnormalities in 80 (39 %), minimal hematuria or proteinuria in 84 (41 %), and severe proteinuria or hematuria in 40 (20 %). Many children in the third group suffered multiple recurrences of nephritis for more than 1 year. Only four cases (2 %) showed progressive deterioration in renal function. HSP has been reported in adults with both higher and similar frequencies of renal disease as in children with HSP. A retrospective analysis of 162 children and adults with HSP showed more frequent and severe renal involvement in adults.
The extent of hematuria, proteinuria, and renal dysfunction at presentation has been shown to correlate with greater morphologic severity or glomerular disease, which in turn has been associated with worse long-term renal outcomes. Despite these clinicopathologic correlations, extended follow-up has revealed unexpected deterioration in several HSP patients who had only mild renal manifestations at the initial assessment. Furthermore, the predictive value of an initial biopsy is poor because many patients with severe glomerular changes recover completely from the illness. For these reasons, a renal biopsy is not recommended in the routine initial evaluation of HSP patients.
Treatment Studies
The evidence to date does not show that immunosuppressive therapy alters the natural history of HSP. Glucocorticoids play only a limited role in the treatment of HSP. Adults require drug therapy more frequently than do children. In the acute illness, glucocorticoids can promptly relieve severe abdominal pain and arrest gastrointestinal hemorrhage. Glucocorticoids and cytotoxic agents have been prescribed during the acute illness to prevent serious renal disease, but they do not appear to change long-term outcomes. An uncontrolled study of low-dose immunoglobulin therapy has also been performed in patients with HSP and renal disease, but the outcomes, although favorable, are difficult to interpret without a control group. In contrast, other investigators have reported a case of HSP in which renal function deteriorated after immunoglobulin therapy.
In a prospective trial, 168 children with HSP who lacked renal abnormalities at presentation were assigned to receive either no treatment of 1 mg / kg / day of prednisone for 2 weeks. None of the prednisone treated patients developed clinical signs of nephropathy. Microscopic hematuria, proteinuria, or both appeared within 2 to 6 weeks after recovery from the acute illness in 10 of the untreated patients. These abnormalities resolved within 1 year in 8 of these children, whereas in the other 2 children the urinary findings persisted without decline in renal function. Although this study suggests that prednisone therapy may reduce the incidence of early HSP renal disease, it does not address the critical question of whether glucocorticoids prevent progression to chronic renal failure. 
Patients with HSP and severe nephritis have usually been treated with aggressive glucocorticoid therapy, often in combination with other immunosuppressive drugs, Twelve children aged 6 to 14 years with rapidly progressive glomerulonephritis were treated with pulse methylprednisolone for 3 days, followed by oral prednisolone for 3 months, oral cyclophosphamide for 2 months, and oral dipyridamole for 6 months. The serum creatinine normalized in all but one patient by the end of therapy. Seven patients achieved a complete remission with no urinary sediment abnormalities. The remaining patient developed chronic renal insufficiency with continued hematuria and proteinuria. Other studies have shown regression of glomerulonephritis with intensive immunosuppressive therapy.
Most children and adults with HSP can be managed conservatively as outpatients during the acute illness. Antibiotics may be necessary to treat an infection. Salicylates and other nonsteroidal anti-inflammatory drugs (NSAIDs) can alleviate symptoms from purpura and joint manifestations. Patients with severe abdominal pain or other gastrointestinal complications require hospitalization. Severe abdominal pain, melena, or hematemesis usually responds promptly to 1 mg / kg / day of prednisone, which should not be continued longer than 2 weeks. Persistence of abdominal pain beyond 24 hours or a tender abdominal mass may be a sign of intussusception. Such findings mandate further diagnostic evaluation and surgical consultation. 
Patients with rapidly progressive glomerulonephritis should be treated with high dosages of glucocorticoids in combination with cyclophosphamide. Treatments should be initiated with IV methylprednisolone, 0.5 mg / kg / day for 3 days, and oral cyclophosphamide, 2 mg / kg / day. After the methylprednisolone pulses, the patient is switched to oral prednisone, 1 mg / kg / day for 1 month, followed by tapering of the dose as described earlier. The appropriate duration of oral cyclophosphamide therapy is unclear, but it probably should be continued for at least 3 to 6 months, depending on the treatment response.
The term hypersensitivity vasculitis defines a heterogeneous group of vasculitic disorders in which the predominant pathologic lesion is inflammation of arterioles, capillaries, and venules. Although virtually any organ system or vessel size can be attacked in this condition, skin manifestations and biopsy evidence of leukocytoclastic vasculitis dominate the clinical and histopathologic picture. The ACR has developed criteria for the classification of hypersensitivity vasculitis. The cutaneous disease can variously appear as purpura, urticaria-like lesions, bullous lesions, ulcers, erythematous plaques, or nodules. Less frequently, hypersensitivity vasculitis may produce vessel inflammation outside the skin, leading to joint, kidney, liver, lung, heart, or nervous system involvement.
As the name implies, hypersensitivity vasculitis is thought to represent an immune response to an antigenic stimulus resulting from exposure to an environmental toxin, foreign protein, drug tumor antigen, infectious agent, or self-antigen. Historically, several clinical syndromes have been included under the general heading of hypersensitivity vasculitis: drug induced vasculitis, serum sickness, HSP, type II MC, vasculitis associated with other primary disorders. Hypersensitivity vasculitis is a diagnosis usually reserved for patients with a small-vessel vasculitis apart from HSP, type II MC, and the secondary causes of small-vessel vasculitis.
Natural History of Disease
The heterogeneity of hypersensitivity vasculitis as a diagnostic category has obscured our understanding of its natural history. Several subgroups of hypersensitivity vasculitis have been described. Serum sickness occurs after exposure to a foreign antigen (e.g., heterologous horse serum) and typically produces a 2 to 4 week self-limited bout of fever, urticaria, arthralgias, and lymphadenopathy. The clinical course of patients with hypersensitivity vasculitis of unknown cause is more variable than that of serum, sickness. In one study of 82 patients with this condition, the clinical course was characterized as acute (56%), chronic (28%), or relapsing (16%), with recurring episodes of skin disease separated by intervals of months to years. The subgroup of patients with drug-related hypersensitivity vasculitis also has a highly variable course. Often lacking in studies is convincing proof that the drug actually caused the vasculitis.
Treatment Studies
Discontinuation of a potentially inciting drug has been shown to be effective therapy for most cases of drugs. 
Related hypersensitivity vasculitis.  Skin involvement has been reported to respond to treatments with topical glucocorticoids, antihistamines, and NSAIDs 136 for patients with chronic or relapsing disease, Callen 138 reported success with oral colchicines in a dosage of 0.6 mg twice daily.  However, the results of a small, randomized, controlled trial did not find the colchicines, 0.5 mg twice daily, was more effective than topical emollients in preventing relapses of cutaneous leukocytoclastic vasculitis.139  Other agents, including hydroxychloroquine, dapsone, and azathioprine, have also been reported to produce favorable responses in refractory cases.  A retrospective analysis of 95 patients from a university hospital suggest that hypersensitivity vasculitis has an excellent outcome regardless of therapy.  140  Only two of the patients in this study failed to recover completely. 
The first step in treating hypersensitivity vasculitis is removal of the offending antigen.  Because the vasculitic process is often self-limited, it is treated conservatively with local measures, including topical glucocorticoids, protective dressings, and Unna boots, as well as antihistamines and NSAIDs.  Treatment with colchicines, hysdroxychloroquine, dapsone, or azathioprine may be considered in patients with chronic or relapsing disease.  A short course of prednisone (<20mg/day) may be warranted in selected patients with severe skin or joint manifestations.  Although systemic glucocorticoids can suppress the development o new skin lesions, their withdrawal is often followed by the appearance of new eruptions.  Evidence does not support the prolonged use of glucocorticoids or cytotoxic agents for treating hypersensitivity vasculitis.  However, in rare cases, such as severe bullous or ulcerative disease, this approach may be undertaken with caution.  The occasional patient with serous or life-threatening systemic involvement may require treatment with 40 to 60 mg/day of prednisone, followed by a 2-to-3-month taper to the lowest dosage that effectively controls the disease.  
Cogan’s syndrome is a rare disorder characterized by recurrent episodes of bilateral ocular inflammation and vestibuloauditoy dysfunction.141, 142  The hallmark of the ocular involvement is interstitial keratitis, a patchy mononuclear cell infiltrate in the cornea that causes eye discomfort, redness, and photophobia.  Other types of ocular inflammation, including conjunctivitis, kiritis, episcleritis/scleritis, posterior uveitis, vitritis, choroidkeratitis. 141  Vestibuloauditory disease, the other major feature of Cogan’s syndrome, provokes sudden and debilitating attacks resembling Me’nie’re’s disease. These attacks may be characterized by nausea, vomiting, hearing impairment, and severe vertigo. In about half of the cases, onsets of ocular and vestibuloauditory symptoms are simultaneous; in the remainder, the disease unfolds as isolated ocular and vestibuloauditory events occurring within months of each other.143 
Systemic inflammatory disease arises commonly in Cogan’s syndrome and may produce constitutional symptoms, central nervous system manifestations, cutaneous lesions, and systemic necrotizing vasculitits.141, 143, 144   Widespread vasculitis develops in about 105 of patients and can affect either large vessels (resembling Takayasu’s disease) or medium-sized vessels (PAN-like)143  Aortis insufficiency with aortitis has been reported in up to 10% of patients with Cogans syndrome. 144
The ocular outcomes on Cogan’s syndrome are excellent unless the course is complicated by posterior segment inflammation (e.g. slceritis, vitritis, osterior uveitis).  Interstitial keratitis, conjunctivitis, and iritis target the anterior segment of the eye and only rarely cause visual loss.  In contrast, ocular inflammation of the posterior segment, if left unchecked, can rapidly destroy ocular tissue and cause permanent visual loss.  Blindness has been reported in about 5% of patients with Cogan’s syndrome and results from ocular inflammation beyond the anterior segment.  141,144  
Hearing loss, sometimes asymmetric, is the major cause of disability in Cogan’s syndrome.  Deafness was noted in 21 of 78 ears in one series. 143 and in 78 of 156 ears in another previous review.144 Vestibular symptoms usually resolve, although mild oscillopsia may persist in the minority of cases.  Cochlear hydrops can develop as a consequence or prolonged disease and produce hearing fluctuations as a result of hormonal changes (e.g., the menstrual cycle in women), ingestion of salty foods, allergic conditions, or an upper respiratory illness.  Such hearing fluctuations are clinically indistinguishable from those of inflammatory origin. 
Prompt treatment of interstitial keratitis with topical glucocorticoids usually prevents corneal scarring and vascularization. 143 The rare occurrence of severe visual loss from corneal scarring has been treated with corneal transplantation.  Posterior segment ocular disease has responded variably to treatment with systemic glucocorticoids, cyclophosphamide, and cyclosporine. 61,141,145,146
Acute hearing loss may be at least partially reversed by systemic prednisone therapy, 1 to 2 mg/kg/day. 145 Some evidence suggests that prompt therapy with systemic glucocorticoids can reduce hearing loss. 147 Acute hearing loss resistant to glurcocorticoid therapy has been managed successfully with oral cyclophosphamide, methotrexate, and azathioprine therapy 145; however, the true efficacy of these agents is uncertain, in view of the variable nature of the disease.  
Aortic insufficiency resulting from aortitis has been ameliorated after treatment with glucocorticoids and cytotoxic agents. 141  Other patients have required aortic value replacement. 141-144, 148 Large-and-medium-vessel vasculitis in Cogan’s syndrome can be controlled with prednisone, oral cyclophosphamide, and cyclosporine. 61,141,143 Occlusive vascular disease caused by coronary 141 and mesenteric 149 arteritis has necessitated surgical bypass to correct end-organ ischemia. 
Interstitial keratitis and iritis are satisfactorily managed with mydriatics and topical glurcocorticoid therapy.  Ocular signs and symptoms generally improve within 3 to 7 days after the start of therapy.  A poor treatment response may signal the presence of another disease process such as chlamydial infection, which can mimic the ocular manifestations of Cogan’s syndrome. 150 Episcleritis and scleritis also usually resolve with topical glucocorticoid therapy.   In contrast, nodular scleritis, posterior uveitis, and retinal artery disease often mandate systemic therapy, consisting initially of 1 to 2 mg/kg/day of prednisone for 1 month, followed by a gradual dosage taper over the next 2 to 3 months (see Table 27-2).  Serious ocular inflammation that progresses despite glucocorticoid therapy can be treated with oral cyclophosphamide or cycloporine. 146 Because such cytotoxic therapy is potentially hazardour, end points must be set to define the period of therapy.  
Audiograms are obtained routinely to quantify hearing loss and monitor the response to therapy.  Patients with acute hearing loss are initially given a 2-week trial of prednisone at a dosage of 1 mg/kg/day.  This dosage is usually maintained for an additional and tapered within 6 to 8 weeks to an alternate-day regimen as long as auditory acuity is preserved.  Some patients require more continuous low does of prednisone on a daily or alternate-day schedule to prevent further deterioration in hearing, whereas others can stop taking prednisone for variable periods and maintain their hearing.  Patients whose auditory acuity does not improve with glucocorticoids may be treated with oral cyclophosphamide (1 to 2 mg/kg/day), methotrexate (7.5 to 20mg/week), or azathioprine (2mg/kg/day).  Because the benefits of these agents are uncertain, therapeutic response criteria are mandatory for defining the period of treatment.  Patients with complete deafness may benefit from a cochlear implant. 
The heart should be periodically examined for signs of aortic insufficiency.  Cardiac symptoms, signs of heart failure, or clinical evidence of a valvular abnormality merits further diagnostic evaluation with two-dimensional echocardiography or cardiac catheterization.  The treatment of systemic inflammatory vascular disease in Cogan’s syndrome parallels that described for PAN. 
TABLE 27-4 
Guidelines for Initial Therapy of the Major Vasculitic Syndromes 
Initial Remission Rate 
Disease Category Initial Therapy with Standard Therapy
PAN, MPA, and CSS Mild disease: prednisone    More than 80% 
Serious disease: prednisone
1mg/kg/d, + oral 
1-2 mg/kg/d
Hepatitis B virus-related Interferon-alpha, 5 MU/d
Wegener’s granulomatosis Severe disease: prednisone 75%
Limited disease: prednisone
1mg/kg/d, +methotrexate,
Hepatitis C virus-related   Interferon-alpha,3 million U thrice 60%-70%
Type II mixed weekly, +short course of 
Prednisone (optional) + plasma
Exchange (optional)
CSS= Churg-Strauss syndrome; MPA= microscopic polyarteritis; PAN = polyarteritis nodosa
The management of the major vasculitic syndromes is summarized in Table 27-4.  These recommendations serve as treatments guides; it is recognized that the standard approach can entail modification in individual cases.  Beyond current therapy, future progress in the management of vasculitis will probably result from new insights into disease pathogenesis. 
The author thanks David Pisetsky, Rex McCallum, Nancy Allen, and Barton Haynes for reviewing this chapter and