Clinical implication of plasma exchange on life-threatening antineutrophil cytoplasmic antibody-associated vasculitis

Antineutrophil cytoplasmic antibody (ANCA)-associated vasculitis (AAV) is one of the small vessel vasculitides and comprises microscopic polyangiitis (MPA), granulomatosis with polyangiitis (GPA) and eosinophilic granulomatosis with polyangiitis (EGPA) [1, 2]. The latest recommendations for the management of AAV were proposed by the European League Against Rheumatism (EULAR) and the European Renal Association-European Dialysis and Transplant Association (ERA-EDTA) in 2016 (the 2016 EULAR/ERA-EDTA recommendations) [3]. According to these recommendations, a combination of glucocorticoid with either cyclophosphamide (CYC) or rituximab (RTX) should be administered to patients newly diagnosed with life-threatening AAV. Once remission is achieved, azathioprine, methotrexate or RTX could be maintained and glucocorticoid may be tapered. In particular, in cases of renal failure with rapidly progressive glomerulonephritis (RPGN) or diffuse alveolar haemorrhage (DAH), plasma exchange (PLEX) should be considered [3]. In addition to the 2016 EULAR/ERA-EDTA recommendations, the 2012 Kidney Disease: Improving Global Outcomes (KDIGO) clinical practice guideline for glomerulonephritis recommends PLEX for patients with rapid and severe renal vasculitis and suspected anti-glomerular basement membrane (anti-GBM) glomerulonephritis [4]. Therefore, PLEX is not usually prescribed but is strongly recommended for patients with life-threatening AAV.

With regard to RPGN, the Methylprednisolone versus Plasma Exchange (MEPEX) clinical trial compared the efficacy of PLEX and methylprednisolone pulse therapy in 137 AAV patients with serum creatinine > 5.8 mg/dL and reported that PLEX for induction exhibited a higher frequency of renal recovery [5]. In addition, another clinical trial reported that PLEX had a preventive efficacy against the exacerbation of renal dysfunction in AAV patients with serum creatinine < 5.7 mg/dL in GPA patients [6]. In contrast, a long-term observational clinical study, PLEX was found to have no significant benefit in AAV patients with serum creatinine > 5.7 mg/dL or on dialysis [7]. Moreover, recently, the data from the Plasma Exchange and Glucocorticoids for Treatment of Anti-Neutrophil Cytoplasm Antibody (ANCA)-Associated Vasculitis (PEXIVAS) clinical trial reported that PLEX had no efficacy for reducing the risk of both end-stage renal disease or death in 704 AAV patients [8]. With regard to DAH, several case series reported the positive efficacy of PLEX on DAH [9, 10]. In particular, a retrospective study including 20 AAV patients with DAH reported that combination therapy with PLEX and immunosuppressive drugs was beneficial for lung outcomes [11]. In contrast, a retrospective cohort study including 53 AAV patients with severe alveolar haemorrhage reported no significant difference in mortality between the PLEX and non-PLEX groups [12]. Therefore, the efficacy of PLEX in life-threatening AAV remains controversial.

To date, there has been no study investigating the efficacy of PLEX on life-threatening AAV, such as RPGN and DAH, on a large number of patients, except for a few case reports from Korea. Hence, in this study, we investigated two clinical implications of PLEX on life-threatening AAV. First, we assessed the rate of and predictors for all-cause mortality in AAV patients receiving PLEX. Second, we assessed the survival-benefit of PLEX for DAH between AAV patients receiving PLEX and those not receiving PLEX.

In this study, we arrived at two conclusions regarding the efficacy of PLEX on life-threatening AAV. Firstly, in terms of the rate of and predictors for all-cause mortality in AAV patients receiving PLEX, the rate of all-cause mortality was found to be 44.4%; however, no significant predictor for all-cause mortality was determined. In the MEPEX trial, the rates of all-cause mortality at 3 and 12 months were 16 and 27%, respectively, in the PLEX group, which is reflective of renal involvement as a serious manifestation in combination with a high risk of infection owing to immunosuppressive therapy [5]. Meanwhile, the PEXIVAS trial included two groups: 352 patients in the PLEX group and 352 in the no PLEX group based on glucocorticoid therapy. The rate of all-cause mortality and end-stage renal disease (ESRD) occurrence was 28.4% in the PLEX group and 31.0% in the no PLEX group. The HR of PLEX on all-cause mortality compared to that of no PLEX was 0.87 (95% confidence interval 0.58–1.29). Therefore, PLEX did not have any influence on the rate of all-cause mortality or ESRD occurrence in AAV patients [8]. It could be assumed that the very high mortality rate might interfere and offset the statistical significance of predictors of all-cause mortality after performing PLEX. In addition, this may have two clinical meanings: first, the therapeutic efficacy of PLEX might not be as high as was expected. Second, the severity of AAV might exceed the therapeutic potential of PLEX on AAV.

Secondly, we compared the survival-benefit of PLEX for DAH between patients with DAH receiving PLEX and those not receiving PLEX. Most previous studies on the efficacy of PLEX were conducted in AAV patients with kidney involvement. Two representative clinical trials, such as MEPEX and PEXIVAS, reported the conflicting efficacy of PLEX on RPGN [5, 8]. As for AAV patients with reduced kidney function due to RPGN, there are alternative treatment-modalities in addition to PLEX: transient renal replacement together with combination therapy comprising a high dose glucocorticoid and either CYC or RTX may improve RPGN [3, 4]. However, as for patients with DAH, there is no alternative treatment-modality in addition to PLEX because DAH is more rapidly progressive and fatal than RPGN. Therefore, it seems impossible to design and conduct randomised case-controlled clinical trials in AAV patients with urgent DAH, unlike that with RPGN.

There was an observational case series investigating the efficacy of PLEX on DAH in 12 AAV patients who were admitted to ICU. The authors demonstrated that PLEX together with a combination of glucocorticoid and immunosuppressive drugs might have a benefit to improve both the respiratory dysfunction and AAV-related DAH in AAV patients, although one patient died. In this study, 10 AAV patients with DAH, of whom six patients received PLEX, were retrospectively analysed [10]. However, unlike this previous study, our current study did not find any efficacy of PLEX for DAH with regard to the improvement of all-cause mortality. We considered the follow-up period as the reason for this discrepancy. The previous study evaluated the SpO2/FiO2 ratio and assessed the mechanical ventilation mode hourly for 7 days. However, they did not evaluate all-cause mortality after extubation and during follow-up [10]. Thus, it could not be easily accepted that PLEX is beneficial for the improvement in the prognosis of AAV-related DAH. In contrast, the median follow-up duration of our study was significantly longer than that of the previous study (1145.5 days for four patients not receiving PLEX and 130.0 days for six patients receiving PLEX). Of the four deceased AAV patients with DAH, three patients died of sepsis due to secondary pneumonia and one patient died because of the rapid progression of DAH owing to the ineffectiveness of PLEX (Table 3).

Considering the results of both our study and previous studies, we would suggest the use of therapeutic strategies depending on the time-course. Firstly, within 1 or 2 weeks from DAH development, PLEX along with combination comprising glucocorticoid with CYC or RTX should be promptly initiated under close observation in the intensive care unit. This strategy is expected to reduce the rate of all-cause mortality from 8.3 to 10.0%. Next, after 2 weeks from DAH development, the most common reason for death was secondary pneumonia and sepsis. Therefore, the monitoring of pneumonia occurrence, the use of preventive antibiotics and the choice of dose or type of immunosuppressive drugs should be carefully considered.

Our study has several limitations. Firstly, we provided a narrative report rather than an analytical one as the number of patients who received PLEX was too small to conduct a subgroup analysis. In particular, for this reason, we realise that the reliability of the Cox proportional analysis of this study might be limited and low. Secondly, we could not obtain sufficient information from the medical record such as PLEX technique, ventilator modes and lung-involvement pattern owing to the limitation of the retrospective study design. Thirdly, this study analysed data that had been accumulated over 8 years. During this period, the considerable improvement has taken place not only in the overall prognosis of AAV patients but also in the accuracy of ANCA tests and the efficacy of immunosuppressive drugs. Therefore, this might influence the results of this retrospective and observational study. Despite these limitations, we believe that our study, which is based on the largest cohort in Korea, could provide valuable information on PLEX in Korean patients with AAV. Therefore, this study has clinical significance as a pilot study.

Nine of 212 AAV patients (4.3%) received PLEX and the rate of all-cause mortality in nine AAV patients receiving PLEX was found to be 44.4%. DAH was the most common reason for using PLEX; however, the notion that PLEX is beneficial for the improvement in the prognosis of AAV-related DAH was deemed controversial.