Disclosures Learning objectives Introduction Granulomatosis with polyangiitis (GPA) is one of a group of primary systemic small vessel vasculitides characterized by the presence of anti-neutrophil cytoplasmic antibodies (ANCA) and referred to, therefore, as the ANCA-associated vasculitides (AAV). Patients with GPA are generally grouped into one of two broad categories: limited (or non-severe), and severe. Patients with limited disease tend to have primarily granulomatous inflammation, which is generally confined to the respiratory tract. Patients with severe, or generalized, disease have manifestations that impact life, or the function of a vital organ, generally as the result of small vessel vasculitis. Such severe disease may present with life-threatening manifestations such as a rapidly progressive pauci-immune glomerulonephritis or diffuse alveolar hemorrhage. Here we focus on key milestones in the evolving management of severe disease, and the basis for an emerging view that long-term remission without steroids or other immunosuppressive medication may be possible for some patients. The Fauci–Wolff protocol: The first remission induction strategy Nearly 45 years have passed since Anthony Fauci and Sheldon Wolff proposed the first effective remission induction strategy for generalized, or severe, GPA . In the absence of immunosuppressive therapy, patients had a mean survival of 5 months, and an equally dismal 1-year mortality rate of 82%. Glucocorticoid therapy alone improved the mean survival time to approximately 12 months. Concomitant use of the alkylating cytotoxic agent cyclophosphamide (CYC), in what is now known as the Fauci–Wolff protocol, dramatically improved patient outcomes. With this protocol, over 90% of patients experience clinical improvement, and 75% enter clinical remission . However, relapse rates after induction with either oral or intravenous CYC were also high, emphasizing the need for alternative long-term immunosuppressive therapy to maintain remission. Relapse rates after withdrawal of CYC were as high as 60% [2, 3]. Additionally, the toxicity of CYC precludes its use for the long-term maintenance of remission. In the longitudinal review of outcomes in the original series of patients treated with the Fauci–Wolff protocol, Hoffman and colleagues reported a sobering picture, with roughly 40% of patients afflicted with some permanent morbidity due to treatment side effects . Nearly half of patients in the protocol suffered from severe infections. Also, nearly half of all patients developed hemorrhagic cystitis, and the risk of bladder cancer and lymphoma were increased by over 30-fold and 10-fold, respectively. Almost 60% of patients in this series became infertile. Steroid-induced damage superimposed on the effects of CYC toxicity was also significant. Remission maintenance with azathioprine As a result, almost all clinical investigations since that time have been guided by a desire to minimize the use of CYC, if not avoid its use altogether. The seminal CYCAZAREM trial published in 2003 validated the now standard concept of a sequential treatment strategy, beginning with a short-term, high-intensity remission induction agent, followed by the use of a less toxic, longer-term remission maintenance agent . In this non-inferiority trial, Jayne and colleagues concluded that inducing patients with severe AAV with CYC for at least 3 months followed by transition to azathioprine 2 mg/kg/day for a total of 18 months of treatment, versus continued treatment with CYC (at 1 mg/kg/d) for 18 months, resulted in similar relapse rates but lower rates of renal morbidity and cumulative damage. Long-term follow-up data subsequently confirmed that, with respect to severe renal manifestations, remission maintenance with azathioprine was non-inferior to remission maintenance with azathioprine with respect to severe renal manifestations; that said, patients who received remission maintenance with azathioprine experienced a threefold increase in non-severe disease relapse, generally affecting the respiratory tract . Regardless, the feasibility of a two-step approach of high intensity induction followed by consolidation with a milder maintenance agent was reaffirmed with respect to potential life-threatening complications. The introduction of rituximab After CYCAZAREM, the third and most recent major revolution in the management of GPA has been the introduction of rituximab (RTX) into the therapeutic armamentarium [6, 7, 8]. In 2010, the American-led RAVE and the European-led RITUXIVAS non-inferiority trials independently demonstrated comparable remission induction rates in patients treated with either CYC or RTX; the RAVE trial demonstrated that RTX might be superior to CYC for remission induction among patients with relapsing disease. Furthermore, in the long-term follow up of subjects enrolled in RAVE, published in 2013, Specks and colleagues observed that a single induction course of RTX given in 4 sessions, 1 week apart, was comparable in efficacy to a standard CYC-azathioprine induction course as established by Jayne and colleagues approximately 10 years previously. In addition, the benefits of RTX administered in this manner were achieved with a significantly lower adverse event rate and overall toxicity profile compared with the conventional CYC-azathioprine combination. The logical question arose as to whether or not RTX itself could be used safely and effectively not just for remission induction, but also for remission maintenance. Indeed, in the subsequent MAINRITSAN trial published in 2014, Guillevin and the French Vasculitis Group found that low dose RTX administered in 6-month intervals was, in fact, greatly superior to azathioprine in maintaining remission in GPA patients up to at least 28 months after completing remission induction with CYC . Differences in relapse rates in the azathioprine versus RTX arms were striking, at 29% and 5% respectively. Is drug-free remission possible? We and others believe that RAVE and MAINRITSAN suggest that eventual discontinuation of all immunosuppression under careful monitoring may be possible for some patients with generalized GPA. In the recently concluded MAINRITSAN 2 study, the French Vasculitis Group found similar efficacy of scheduled RTX redosing every 6 months versus intermittent redosing, triggered by the recrudescence of CD19 and ANCA in the serum . Relapse rates were notably low at roughly 15% for both study arms despite an average of 40% fewer RTX doses in the tailored arm of the study. This observation has led to MAINRITSAN 3, which proposes to extend the findings of MAINRITSAN 2 further by comparing outcomes when RTX is stopped altogether versus continued according to a 6-month redosing schedule across an additional 28 months beyond the original 28 month observation period in MAINRITSAN 2. Notably, one important caveat of RITUXIVAS, MAINRITSAN, CYCAZAREM and almost all other GPA clinical trials has been allowance within the definition of “drug-free” remission maintenance of long-term use of a presumed negligible dose (5 mg or less) of daily prednisone. We and several others have wondered about the possibility of pushing even these limits to see if sustained remission that is truly drug-free is actually possible for some patients [11, 12, 13]. We recently began examining this issue specifically in our own center in a recent case series of 18 patients with generalized AAV induced either with CYC or RTX who were either contraindicated for standard maintenance immunosuppression or expressed a strong informed preference to avoid indefinite long-term medication use. Interestingly, we found that disease-free remission was actually possible in over half of these patients . Prednisone was weaned completely in 12 patients, with the median duration of prednisone use among these 12 patients being 20 months (range 6–42 months). A meta-analysis of 13 studies of AAV (representing 983 patients) indicates that faster steroid tapers are associated with an increased risk of relapse . A formal controlled trial known as TAPIR and sponsored by the Vasculitis Clinical Research Consortium is currently enrolling to directly compare outcomes in GPA patients in remission who will be randomized to receive either low dose of prednisone or no prednisone at all. Conclusion To conclude, total withdrawal of all steroids and immunosuppressive agents almost certainly will not be possible for all patients with generalized GPA. Still, we believe that clinicians one day in the near future will have access to standardized guidelines for predicting which patients can tolerate even reduced intensity long-term maintenance therapy versus those who can stop altogether and remain in remission under careful monitoring. Our field looks forward to this forthcoming fourth revolution in the management of GPA.