including 230 EGPA patients [42]. The revised FFS retained the following 5 factors associated with a higher mortality risk rate: serum creatinine >150 μmol/L, severe gastrointestinal involvement, myocardial involvement, age > 65 years, and the absence of ENT manifestations, each accorded 1 point; then, the points were added. Adding the points yielded the FFS; scores = 0, 1, or 2 were associated with respective 5-year mortality rates of 9%, 21%, or 40%. Hence, calculation of the revised FFS with these readily available parameters can help identify vasculitis patients at a high risk of death who require aggressive therapy with immunosuppressants. In our series, 11.7% of the patients died at a mean ± SD of 50.4 ± 60.1 (median 21.4) months post-diagnosis. The leading cause of death was heart-related (31%; i.e., myocardial infarction, cardiac insuf ciency, or arrhythmia), followed by infections or malignancies (11% each), active vasculitis, and respiratory failure (severe asthma attacks and/ or end-stage chronic pulmonary obstructive disease, 9% each) [23].

Vasculitis relapses remain a major concern for EGPA. They are de ned as the new appearance, recurrence, or worsening of clinical EGPA vasculitis signs (excluding asthma and/or ENT) that require adjunction, change, or intensi cation of the glucocorticoid and/or other immunosuppressant dose(s) [60]. Although almost 90% of EGPA patients achieve remission, 25.3% of those in our series relapsed and 18% experienced asthma fares, sinusitis, and/ or enhanced eosinophilia levels, justifying the prolonged use of glucocorticoids for about 85% of patients in the series [23]. The relapse-free survival rate was 78.6% (95% CI 64.3–84.3) for all patients, and most relapses occurred during the rst year post-diagnosis.

Phenotypes According to the ANCA Status

Depending on the ANCA status, EGPA clinical ndings and prognoses differ markedly, de ning two distinct phenotypes [61, 62].

Based on a series of 93 patients, Sinico et al. showed that ANCA positivity was associated with more frequent ­pulmonary hemorrhages (20% vs. 0%, P = 0.001), renal involvement (51.4% vs. 12.1%, P < 0.001), mononeuritis multiplex, and purpura but less frequent lung in ltrates (34.2% vs. 60.3%) and cardiac involvement (5.7% vs. 22.4%, P = 0.042) [61]. Similar ndings were reported for another series of 112 EGPA patients, with ANCA positivity associated with more frequent renal involvement (35% vs. 4%), peripheral neuropathy (84% vs. 65%), and biopsy-proven vasculitis (79% vs. 39%) [62]. Thus, ANCA positivity might be associated with more frequent vasculitis-associated symptoms, whereas ANCA negativity would be associated with more eosinophilic in ltrates. Patients’ prognoses also differ

according to the ANCA phenotype: in our series, ANCA positivity was associated with more frequent relapses (35.2% vs. 22.5%, P = 0.01) but fewer deaths (5.6% vs. 12.5%, P < 0.5) [23]. These events could be representative of two distinct EGPA phenotypes, de ned by ANCA positivity and distinct genetic pro les.

However, in a series of 157 patients, de nite features of vasculitis were found in 28% of patients who had no ANCAs, suggesting that ANCAs per se are not suf cient to dichotomize patients with or without vasculitis features [63]. Some authors suggested a more precise nomenclature, distinguishing EGPA, with de nite vasculitis features, de nite surrogates of vasculitis, mononeuritis, and/or ANCAs and any systemic manifestation, from hypereosinophilic asthma with systemic manifestations [63].

Treatment

Therapeutic Strategies

EGPA patients can bene t from glucocorticoids, immunosuppressants, and now targeted therapies. As for other AAVs, the treatment strategy for EGPA comprises remission induction and maintenance phases. However, treating EGPA can be dif cult because of asthma and/or ENT manifestations, which can persist even after vasculitis remission has been obtained. In 2015, the EGPA Consensus Task Force published recommendations for the evaluation and management of EGPA patients [59].

In our opinion, each patient should receive a tailored regimen, adapted to different parameters, e.g., disease severity, organ involvement, prognosis, and comorbidities. While the FFS was devised to assess prognosis, using this score to guide the choice of therapeutic regimen remains debated; international recommendations also remain contradictory as to whether to do so [64].

For patients with a 1996 FFS = 0, i.e., without poor prognosis factors, we recommend starting glucocorticoids alone, as their ef cacy to induce and maintain remissions and safety have been demonstrated. In the prospective randomized CHUSPAN trial that enrolled 72 EGPA patients without poor prognosis factors, 93% achieved remission with glucocorticoids alone and 97% achieved a 5-year survival rate [63]. However, 35% of them relapsed, mainly during the rst year of treatment, and, eventually, immunosuppressant adjunction was needed to limit relapses. Again, controversy persists as to immunosuppressant ef cacy for those EGPA patients. In the randomized, prospective CHUSPAN2 trial, azathioprine adjunction to glucocorticoids did not impact the remission or relapse rate [65]. In addition, no steroid-sparing was achieved with azathioprine. The recent European League Against Rheumatism/European Renal Association–European

Dialysis and Transplant Association (EULAR/ERA-EDTA) recommendations have not strati ed treatment guidelines according to the FFS, despite the results of several prospective trials clearly highlighting its pertinence [66]. Other immunosuppressants, e.g., methotrexate or mycophenolate mofetil, could be other therapeutic options, but prospective, controlled trial ndings did not demonstrate the ef cacy of one or the other and results obtained with azathioprine were better.

For patients with FFS ≥ 1, i.e., with poor prognosis factors, and/or with other lifethreatening manifestations, including those not part of the FFS (i.e., possible blindness due to eye involvement, severe alveolar hemorrhage, and/or fulminant mononeuritis multiplex), it is our opinion that an immunosuppressant should be added to glucocorticoids [59].

Remission Induction

The induction-phase goal is to achieve, as rapidly as possible, clinical remission, de ned as the absence of clinical vasculitis manifestations. Glucocorticoids represent the cornerstone of induction regimens. The initial dose of prednisone (or its equivalent of methylprednisolone) should be ~1 mg/kg/day for 2–3 weeks, followed by gradual tapering. When life-threatening symptoms are present, intravenous (IV) methylprednisolone pulses should be administered (usually 7.5–15 mg/kg/day for 3 days) because of their rapid action. A rapid therapeutic response is usually obtained, with attenuation of asthma and vasculitis symptoms, along with eosinophilia levels. No consensus has been published for glucocorticoid tapering, but the dose should be decreased extremely gradually to the lowest effective dose until withdrawal, when possible. Nevertheless, patients often require long-term glucocorticoids because of steroid-dependent asthma. A daily prednisone dose <7.5 mg could be targeted to reduce glucocorticoid-related adverse events (AEs), but, even at low doses, AEs may occur, and steroid toxicity is cumulative.

When immunosuppressant adjunction is deemed necessary, cyclophosphamide has been the most investigated. In a prospective study on 48 EGPA patients with poor prognosis factors comparing 12 vs. 6 pulses of cyclophosphamide, both with glucocorticoids but with no subsequent maintenance therapy, no statistically signi cant differences were found between the two regimens. Although the complete remission rate for all patients was 78.5% [67], relapses rates remained too high (28.6% vs. 66.7% for 12-versus 6-pulse groups), suggesting the necessity of prescribing maintenance therapy. In an analogy to the induction regimen prescribed for GPA and MPA in our 1997 prospective trial [68], IV pulses or continuous oral cyclophosphamide (2 mg/kg/day) were con-

sidered equally effective, as subsequently con rmed by the CYCLOPS trial [69]. However, oral intake has been associated with higher cumulative cyclophosphamide doses and more severe leukopenia. A post hoc, long-term, follow-up analysis showed that although IV pulses were associated with a higher risk of relapse (39.5% vs. 29.8%, P = 0.029), the mortality rate was not higher [70]. In addition, IV pulses might have the potential advantage of favoring compliance. By analogy with other AAVs, the standard of care has since become 6 cyclophosphamide pulses for induction (3 infusions at 0.6 g/m2 or 15 mg/kg every 2 weeks, followed by 3–6 additional pulses at 15 mg/kg or 0.7 g/m2 every 3 weeks), followed by maintenance therapy [59].

The prospective, randomized, CORTAGE trial evaluated the impact of xed, low-dose, IV cyclophosphamide pulses (500 mg/m2 every 2–3 weeks until remission) versus the standard regimen with limited glucocorticoid exposure on 108 AAV patients ≥65 years old (including 14 with EGPA) [71]. Although the two arms were comparable for remission rates, the patients prescribed the “lighter” regimen suffered fewer serious AEs.

Maintenance Therapy

After remission, a maintenance regimen should be prescribed for poor prognosis patients with an FFS ≥ 1, as they are at an extremely high risk of relapse (73.8–85.7%) [67]. It should start 2 weeks after the last induction cyclophosphamide pulse or a few days after stopping oral cyclophosphamide.

Although many immunosuppressants have been prescribed for EGPA remission maintenance, unlike GPA or MPA, no randomized trial has compared them to identify the more effective agent. As for GPA and MPA, azathioprine can be prescribed, initially at 2 mg/kg/day, but its dose can be adjusted to the clinical response and toxicity. Because mycophenolate mofetil has been shown to be less effective than azathioprine in GPA and MPA remission maintenance therapy, it is not prescribed very often [72]. Finally, based on the results of a prospective trial for GPA and MPA remission maintenance therapy, methotrexate, at an initial dose of 0.25 mg/kg/week, seems to be an acceptable alternative to azathioprine [73].

Although the optimal maintenance therapy duration has not yet been determined, it should last for at least 18–24 months after remission. No consensus criteria to completely stop treatment have been formulated to date. However, according to the long-term follow-up of 118 EGPA patients, relapses occurred at a mean of 2 years post-induction, i.e., when immunosuppression had been terminated for almost all patients, again suggesting the potential long-term need for immunosuppressive therapy [74].