To the Editor:
Eosinophilic granulomatosis with polyangiitis (EGPA) is a systemic anti-neutrophil cytoplasmic antibodies (ANCA)-associated vasculitis affecting the small- and medium-sized arteries. It is characterized by bronchial asthma often accompanied by pulmonary infiltrates, chronic sinusitis, nasal polyps, and peripheral blood eosinophilia, and occasionally followed by a multisystem vasculitic phase involving several organs. Neurological involvement occurs in 51%–86%, usually affecting the peripheral nerves. Conversely, the central nervous system (CNS) is seldom affected (< 10%). In these cases, cerebral ischemic infarction (∼80%) and diffuse encephalopathy (∼10%) have been by far the findings most frequently reported. Published data on EGPA with CNS hemorrhage are limited to anecdotal case reports1–10,11–20,21,22,23,24,25,26. We described a case of EGPA presenting with intracerebral hemorrhage (ICH) and did a comprehensive review of the literature. Ethics board approval was not required in accordance with the policy of our hospital.
A 48-year-old man was admitted because of malaise, intermittent low-grade fever, and mild weight loss in the previous 3–4 months. One month before admission he developed numbness with paranesthesia and burning pain in the distal third of his legs. Over the preceding 18 months he referred to a persistent bronchial asthma. He denied a history of traditional cardiovascular risk factors, drug abuse, and trauma. The blood test showed leukocytosis (20.070/mm3), eosinophilia (8.110/mm3), C-reactive protein (CRP; 49 mg/l), erythrocyte sedimentation rate (ESR; 52 mm/h), and anti-myeloperoxidase (anti-MPO) antibodies (128 U/ml). Coagulation variables and an echocardiogram were normal. An electroneurophysiological study showed a severe axonal mixed sensitive-motor polyneuropathy in the lower limbs. EGPA was diagnosed and a 3-day course of intravenous pulses of methylprednisolone (1 g/day) was initiated. On the third day, he suddenly went into a coma and presented a left-sided hemiparesis. A head computed tomography (CT) scan disclosed a large intracranial right-sided frontoparieto-temporal hematoma with midline shift. A CT angiogram excluded vascular malformations and arterial aneurysms. Biweekly intravenous pulses of cyclophosphamide (500 mg) and oral prednisone (30 mg/day) were started. A month later he partially recovered motility, polyneuropathy completely disappeared, and the size of the hematoma was reduced, but dysarthria remained.
Comprehensive research of the literature on CNS hemorrhage in EGPA was done [MeSH: (Churg-Strauss vasculitis or syndrome) or (eosinophilic granulomatosis with polyangiitis) and (intracerebral, cerebral, intracranial, spinal, subarachnoid, brain} and (hemorrhage, bleeding, hematoma)]. Finally, 27 case reports were considered suitable for our analysis. Also, our case was added to the series. The description of the 28 cases is summarized in Table 1 and Table 2.
No differences in sex were observed. The mean age at bleeding presentation in our series (47 ± 11 yrs) was significantly lower than the mean age of the first event of ICH in the general population (65 ± 12 yrs in men, 69.5 ± 11 yrs in women)27. Virtually all patients had systemic low-grade inflammation characterized by increased ESR and/or CRP, leukocytosis, and hypereosinophilia. In addition, the frequency of ANCA-positivity in our series was found to be ∼2-fold higher (78%) than expected in patients with general EGPA (∼40%). These results support the hypothesis that CNS hemorrhage in EGPA may substantially have a vasculitic origin.
The frequency of hypertension (HTN; usually mild to moderate) in our series was 40%, similar to that reported in patients with general EGPA. Therefore, the contribution of HTN to the CNS bleeding in these patients does not seem to be fundamental from a pathophysiological point of view.
In almost all patients, the CNS hemorrhage was preceded by a well-documented diagnosis of EGPA or a history of typical symptoms, with a highly variable interval between both events. The spectrum of clinical manifestation seen in our patients was comparable to that of the largest series and there was no clinical phenotype associated to CNS bleeding.
Several reports have shown a relatively favorable 5-year survival rate (90%–97%) for patients with EGPA28. In our series, mortality was 15% (4 of 26 cases). Three deaths occurred within 3 months after bleeding and a fourth at 4 years following a recurrence. So, the estimated 5-year survival rate in our series was 85% in the best-case scenario. However, the first 2 deaths occurred before 1995, when the overall prognosis of the EGPA was worse than it is currently28. Therefore, the short-term prognosis of EGPA patients with CNS bleeding is relatively good and possibly comparable to the rest of patients.
The most important limitation derives from the small number of cases analyzed. However, to our knowledge this is the largest series of patients with EGPA affected by CNS bleeding published in the literature14,17.
Despite CNS hemorrhage complicating EGPA being rare (< 1%), the differential diagnosis of patients with CNS bleeding and some of the following characteristics should be taken into account: (1) relatively young patients, (2) absence of an alternative cause to explain the bleeding, (3) a history of EGPA or some of its typical symptoms, (4) coincidence with constitutional symptoms and/or multisystemic involvement, especially peripheral neuropathy, (5) demonstration of blood inflammatory markers and hypereosinophilia, and (6) positivity for ANCA, especially anti-MPO specificity.
REFERENCES
- 1.
- 2.
- 3.
- 4.
- 5.
- 6.
- 7.
- 8.
- 9.
- 10.
- 11.
- 12.
- 13.
- 14.
- 15.
- 16.
- 17.
- 18.
- 19.
- 20.
- 21.
- 22.
- 23.
- 24.
- 25.
- 26.
- 27.
- 28.