Incidence and Temporal Trend in Risk Factors of Severe Infections in ANCA-Glomerulonephritis Patients

HAL is a multi-disciplinary open access archive for the deposit and dissemination of scientific research documents, whether they are published or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers. L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés. Incidence and Temporal Trend in Risk Factors of Severe Infections in ANCA-Glomerulonephritis Patients Pierre Jourdain, Benoit Brilland, Ouassim Medhioub, Jeanne Caron, Clément Samoreau, Assia Djema, Renaud Gansey, Jean-Philippe Coindre, Maud Cousin, Anne Sophie Garnier, et al.

D espite great improvement in anti-neutrophil cytoplasmic antibody (ANCA)Àassociated vasculitis (AAV) management, S1 morbidity and mortality of AAV patients remain significantly higher than in the general population. Indeed, 5-year mortality rate reaches 20% to 25%, 1,2 mainly due to kidney and lung involvement. S2 Causes of death have been widely analyzed, and have related to vasculitis activity, cardiovascular events, cancers, and infections 1,3,S1 in a complex interplay of baseline comorbidities, vasculitis activity, and regimen toxicity. Infections are not only the leading cause of death in the early phases of remission-induction treatment, but also account for substantial mortality in the long term. 4,S1 Thus, the identification of subgroups of AAV patients with a higher risk of infection may be of great help in implementing preventive strategies such as dose regimen adjustment and prophylaxis with anti-infectious drugs.
Previous studies have shown that infection risk is greater within the first months of immunosuppressive treatment, and have suggested older age, renal impairment, pulmonary involvement, high disease activity, steroid dose, or lymphopenia as potential risk factors for infection in AAV patients. 5-8,S3ÀS5 However, to date, no study has addressed whether risk factors of infection vary over time. Thus, the aim of this study was to analyze time-dependent variations in risk factors for severe infections.

Severe Infections and Pathogens
A total of 235 severe infections developed in 90 patients (53.6%), predominantly bacterial (79.6%) and most often affecting the lung (40.6%). Of 105 positive cultures, gram-negative bacteria were the most frequently isolated germs (59%) ( Supplementary Tables S2 and S3). Most patients experienced their first severe infectious episode within the first months following ANCA-GN diagnosis (Figure 1a), and 48 (25%) experienced 3 or more infectious episodes during follow-up ( Figure S1). The overall infection rate was 27.7 episodes per 100 person-years: 140.8/100 person-years between diagnosis and month 3; 60.0/100 person-years between month 3 and month 6, decreasing after month 6, and remaining stable thereafter (Figure 1c and d).

DISCUSSION
To the best of our knowledge, this study is the first to address the risk factors for infection in ANCA-GN patients by performing a temporal trend analysis. The major finding is that risk factors vary over time. Early infections that are associated with the induction immunosuppressive drug regimen are favored by older age and the need for ICU admission at diagnosis. Conversely, late infections are associated mainly with poorer kidney function at 6 months. It is interesting, but not surprising, that kidney impairment was the only significant risk factor associated with late infections in the multivariable analysis, with each 10 ml/min per 1.73 m 2 eGFR decrease conferring a 20% increase in infection risk. This suggests that renal function becomes the main risk factor for infection, more so than other classic risk factors, such as age.
Another important finding was that TMP/SMX prophylaxis was independently associated with a 40% risk reduction of severe infection. Interestingly, this effect was observed mainly for late infections and is in line with a recent study showing a 70% risk reduction of severe infections under TMP/SMX prophylaxis in AAV patients treated with rituximab as a remissioninduction regimen. 9 This may be explained by the fact that besides its action on P. jirovecii, TMP/SMX is also active on bacteria. TMP/SMX is commonly used to prevent P. jirovecii infections and has been recommended by the European League Against Rheumatism (EULAR) and the European Renal AssociationÀEuropean Dialysis and Transplant Association (ERA-EDTA) for patients treated with cyclophosphamide. S6 However, no formal recommendation is given on how long it should be maintained.
We are able to confirm that severe infections are very common in ANCA-GN patients, 2,5 corroborating a recent multicenter retrospective study showing a higher incidence in these patients as compared to an age-and sex-matched population. 2 The high infection risk at ANCA-GN diagnosis may be explained by different commonly admitted factors such as high vasculitis activity and a heavy immunosuppressive drug regimen. Previous studies, not specifically conducted in ANCA-GN patients, have identified older age, impaired kidney function, and dialysis dependency at diagnosis as constant risk factors for infection, 4,5,9 which we also found in our study.
In recent large trials (MEPEX and PEXIVAS trials), S7,S8 infections represented a major risk factor for death in ANCA-GN patients. Our observation that kidney function at month 6 (and not at ANCA-GN diagnosis) is the major risk factor for late infection suggests that infection risk may decrease if kidney function improves following remission-induction treatment. Thus, early ANCA-GN diagnosis and prompt initiation of immunosuppressive treatments, by maximizing the changes in renal recovery, appear as to be major factors in the reduction of infection risk.
It is important, and in line with recent prospective trials, that we did not observe any difference in infection risk according to type of immunosuppressive drug regimen. S9,S10 A previous retrospective study also showed that maintaining a low dose of steroids after 6 months was associated with an increased infection risk. 6 Despite infection risk tending to be higher in patients maintaining steroids at 6 months in our study, it did not reach statistical significance. In opposition to some past studies showing an increased rate of severe infection in patients with earÀnoseÀthroat or pulmonary (i.e., endobronchial stenosis) involvement, 9,S4 we did not observe any impact of extrarenal AAV involvement or of ANCA subtype on infection risk. These discrepancies may be related to the selection of ANCA-GN patients with severe renal impairment in our study.
The limitations of our study must be underlined: the observational and retrospective design, and inclusions covering a 20-year period with substantial modifications in therapeutic strategies. Moreover, vaccination status against influenza and pneumococcus were not available, and their frequencies may have increased over time. Finally, ICU admission was identified as a risk factor for early infection; however we cannot exclude the possibility that infection was in fact the cause of ICU admission at least in some patients.
In conclusion, our study, performed in a large and well-characterized cohort of ANCA-GN patients, S11,S12 displaying an incidence of serious infectious events in line with the current literature, identifies different risk factors over time. Early infections were associated with older age and severity of the vasculitis, or the presence of high comorbidity (ICU requirement), whereas late infections were instead associated with kidney function. Importantly, we found that long-term prophylaxis with TMP/SMX was associated with a decreased risk of mainly late infections. These data open new perspectives for infection prevention in ANCA-GN, suggesting a further focus in prospective studies on the long-term use of TMP/SMX in ANCA-GN patients.

DISCLOSURES
All the authors declare no competing interests.

SUPPLEMENTARY MATERIAL
Supplementary File (PDF) Supplementary References Supplementary Methods Table S1. Baseline characteristics of the population and main outcomes Table S2. Nature of infections observed in the cohort Table S3. Detail of bacteria isolated Table S4. Univariable cox analysis of risk factors associated with first infectious event Figure S1.