Prognosis and ICU Outcome of Systemic Vasculitis
The aim of the current study was to describe the epidemiology, the clinical features, and the outcome of vasculitis patients who were admitted to an ICU. The mortality rate for this population was 52% in the ICU and 58% one year later which was significantly higher than the mortality rate for a general ICU population (less than 20% in our ICU). Our results demonstrate also that SAPS II and BVAS scores were the main independent indicators of prognosis.
Systemic necrotizing vasculitis is a group of diseases defined by vessel wall inflammation that, although heterogeneous, shares some common clinical characteristics and therapeutic management. ICU admissions of such patients are mainly justified by the occurrence of severe active disease related-organ dysfunction or severe complication of immunosuppressive therapy, mostly infection. In this study, severe active manifestation of the disease was the reason of ICU admission in 61% of our patients. A pneumo-renal syndrome with both acute pulmonary and renal failure occurred in 9 patients and, as a manifestation of Granulomatosis with polyangiitis in 8 of them. Cruz and colleagues, including 26 patients with systemic necrotizing vasculitis, found also an active manifestation (77%) as a major reason of ICU. It is noteworthy in their report that vasculitis exacerbation was often the first manifestation of the disease leading to its diagnosis which was the case in 7 (23%) of our patients. In the study of Khan et al., diffuse alveolar hemorrhage was the most frequent reason of ICU admission which is also an active manifestation of the disease. More recently, a severe alveolar haemorrhage was reported as the first disease manifestation in 46 (86.8%) of 53 patients with ANCA-associated vasculitis. Infections represent the second cause of ICU admission as in ours. Eight (26%) patients presented indeed a septic shock requiring critical care management. The well recognized risk of infection in these patients is increased; and is determined primarily by the level of immunosuppression. Actually, all the 8 patients admitted for severe sepsis received more than 20 mg steroids daily at least in the last three months. Godeau et al. showed, in a study including 181 patients with systemic rheumatic disease admitted to ICU that, in the sub-group of 39 vasculitis, the admission infection rate was at 49%. However, the immunosuppressive treatment was not reported precisely.
ICU severity scores and critical care management of our population underlined its high morbidity. More than ¾ of patients were on mechanical ventilation, more than half required vasopressive or inotropic drugs and renal replacement therapy. For comparison, Khan found that 45% of patients were ventilated and 29% underwent hemodialysis. The ICU mortality in our cohort was 52% which is higher than that of a general ICU population. This high ICU mortality rate underscores the pejorative role of the specific disease and/or of long term immunosuppressive treatment. Studies focusing on outcome and prognosis of patients with vasculitis admitted to ICU are scarce. They include mostly systemic rheumatic diseases like connectivitis, rheumatoïd polyarthritis, granulomatosis, primary vasculitis and systemic vasculitis. To the best of our knowledge, only 3 studies investigated specifically vasculitis ICU outcome and reported lower mortality rates varying from 11 to 33,3%. However, these results should be interpreted with caution, at least under the light of the severity of both the specific disease and the associated organ failure. As earlier notified, 80% of our patients were mechanically ventilated with severe respiratory disease and patients with pulmonary vasculitis were reported to have an ICU mortality of more than 50% depending on the disease severity. Last, we observed that two patients among survivors deceased one year after they were discharged from hospital underlying that the occurrence of severe complications requiring an ICU admission may also affect long term survival. Previous studies reported similarly a significant increase in mortality at a one year (from 11 to 29%), and a 30 months (from 15 to 39%) follow-up. In the recent report of Hruskova and colleagues, the long-term prognosis of patients with severe ANCA-associated vasculitis showed that the mortality increased from 17% at 3 months to 41.5% at 49 months median follow-up.
The identification of predictive factors for ICU mortality in such patients is obviously an important matter to study, since it may help in an optimal follow-up and management. Several factors have been previously reported to be predictive of ICU mortality. We identified SOFA and SAPS II severity scores, the need for catecholamine or/and renal replacement therapy and the occurrence of an ARDS as prognosis factors by univariate analysis. However including these parameters in a multivariate analysis, only SAPS II and BVAS scores were associated with an increased risk of ICU mortality. ICU gravity scores especially SAPS II and SOFA scores were previously reported as predictive factors of mortality for vasculitis in ICU in several studies. By contrast, BVAS score was not found to be correlated to ICU mortality. Cruz and colleagues observed similar BVAS scores in survivors and non survivors (16,4±10.3 vs 15.8±20.1; p = 0.75) in a 30- month follow-up. Khan et al., using a modified BVAS/WG score adapted to Granulomatosis with polyangiitis, also failed to correlate it with a 28 day ICU mortality. Frausova and al. reported even a higher, although not significant, BVAS in survivors (16.1 vs 11.5). Our observation need to be confirmed but highly suggests that vasculitis disease activity influence short-term prognosis. Cruz et al. found however that BVAS calculated at ICU admission was significantly higher in non-survivors than survivors at the end of a long-term follow-up of patients with active systemic necrotising vasculitis.
We must acknowledge some limitations to our study. First, it was limited to 4 ICUs from 2 university hospitals and consequently its findings should not be generalized. The sample size is, in addition, relatively small and heterogeneous given the rarity of these diseases. Second, it is a retrospective study which may therefore be affected by various biases. Third, immunosuppressive treatments had changed between individuals which may represent additional bias in the interpretation of the results.
Discussion
The aim of the current study was to describe the epidemiology, the clinical features, and the outcome of vasculitis patients who were admitted to an ICU. The mortality rate for this population was 52% in the ICU and 58% one year later which was significantly higher than the mortality rate for a general ICU population (less than 20% in our ICU). Our results demonstrate also that SAPS II and BVAS scores were the main independent indicators of prognosis.
Systemic necrotizing vasculitis is a group of diseases defined by vessel wall inflammation that, although heterogeneous, shares some common clinical characteristics and therapeutic management. ICU admissions of such patients are mainly justified by the occurrence of severe active disease related-organ dysfunction or severe complication of immunosuppressive therapy, mostly infection. In this study, severe active manifestation of the disease was the reason of ICU admission in 61% of our patients. A pneumo-renal syndrome with both acute pulmonary and renal failure occurred in 9 patients and, as a manifestation of Granulomatosis with polyangiitis in 8 of them. Cruz and colleagues, including 26 patients with systemic necrotizing vasculitis, found also an active manifestation (77%) as a major reason of ICU. It is noteworthy in their report that vasculitis exacerbation was often the first manifestation of the disease leading to its diagnosis which was the case in 7 (23%) of our patients. In the study of Khan et al., diffuse alveolar hemorrhage was the most frequent reason of ICU admission which is also an active manifestation of the disease. More recently, a severe alveolar haemorrhage was reported as the first disease manifestation in 46 (86.8%) of 53 patients with ANCA-associated vasculitis. Infections represent the second cause of ICU admission as in ours. Eight (26%) patients presented indeed a septic shock requiring critical care management. The well recognized risk of infection in these patients is increased; and is determined primarily by the level of immunosuppression. Actually, all the 8 patients admitted for severe sepsis received more than 20 mg steroids daily at least in the last three months. Godeau et al. showed, in a study including 181 patients with systemic rheumatic disease admitted to ICU that, in the sub-group of 39 vasculitis, the admission infection rate was at 49%. However, the immunosuppressive treatment was not reported precisely.
ICU severity scores and critical care management of our population underlined its high morbidity. More than ¾ of patients were on mechanical ventilation, more than half required vasopressive or inotropic drugs and renal replacement therapy. For comparison, Khan found that 45% of patients were ventilated and 29% underwent hemodialysis. The ICU mortality in our cohort was 52% which is higher than that of a general ICU population. This high ICU mortality rate underscores the pejorative role of the specific disease and/or of long term immunosuppressive treatment. Studies focusing on outcome and prognosis of patients with vasculitis admitted to ICU are scarce. They include mostly systemic rheumatic diseases like connectivitis, rheumatoïd polyarthritis, granulomatosis, primary vasculitis and systemic vasculitis. To the best of our knowledge, only 3 studies investigated specifically vasculitis ICU outcome and reported lower mortality rates varying from 11 to 33,3%. However, these results should be interpreted with caution, at least under the light of the severity of both the specific disease and the associated organ failure. As earlier notified, 80% of our patients were mechanically ventilated with severe respiratory disease and patients with pulmonary vasculitis were reported to have an ICU mortality of more than 50% depending on the disease severity. Last, we observed that two patients among survivors deceased one year after they were discharged from hospital underlying that the occurrence of severe complications requiring an ICU admission may also affect long term survival. Previous studies reported similarly a significant increase in mortality at a one year (from 11 to 29%), and a 30 months (from 15 to 39%) follow-up. In the recent report of Hruskova and colleagues, the long-term prognosis of patients with severe ANCA-associated vasculitis showed that the mortality increased from 17% at 3 months to 41.5% at 49 months median follow-up.
The identification of predictive factors for ICU mortality in such patients is obviously an important matter to study, since it may help in an optimal follow-up and management. Several factors have been previously reported to be predictive of ICU mortality. We identified SOFA and SAPS II severity scores, the need for catecholamine or/and renal replacement therapy and the occurrence of an ARDS as prognosis factors by univariate analysis. However including these parameters in a multivariate analysis, only SAPS II and BVAS scores were associated with an increased risk of ICU mortality. ICU gravity scores especially SAPS II and SOFA scores were previously reported as predictive factors of mortality for vasculitis in ICU in several studies. By contrast, BVAS score was not found to be correlated to ICU mortality. Cruz and colleagues observed similar BVAS scores in survivors and non survivors (16,4±10.3 vs 15.8±20.1; p = 0.75) in a 30- month follow-up. Khan et al., using a modified BVAS/WG score adapted to Granulomatosis with polyangiitis, also failed to correlate it with a 28 day ICU mortality. Frausova and al. reported even a higher, although not significant, BVAS in survivors (16.1 vs 11.5). Our observation need to be confirmed but highly suggests that vasculitis disease activity influence short-term prognosis. Cruz et al. found however that BVAS calculated at ICU admission was significantly higher in non-survivors than survivors at the end of a long-term follow-up of patients with active systemic necrotising vasculitis.
We must acknowledge some limitations to our study. First, it was limited to 4 ICUs from 2 university hospitals and consequently its findings should not be generalized. The sample size is, in addition, relatively small and heterogeneous given the rarity of these diseases. Second, it is a retrospective study which may therefore be affected by various biases. Third, immunosuppressive treatments had changed between individuals which may represent additional bias in the interpretation of the results.
SHARE
RELATED POSTS on "Health & Medical"
