A variety of organ systems, such as the kidney, skin, blood vessels and nervous system, often are affected in mixed cryoglobulinemia. In general, immune-mediated vasculitis due to cryoglobulin deposition is considered to be the major pathogenesis (1, 2). In mixed essential cryoglobulinemia, a mild to moderate lung involvement may be seen (3). However, severe lung involvement, leading to life-threatening disturbances of pulmonary function, is very rare in this disorder (3). To our knowledge, only five cases of cryoglobulinemia with severe pulmonary involvement have been described previously. In this report, we describe a case of mixed essential cryoglobulinemia type II due to chronic hepatitis C virus infection with severe lung involvement, observed when the original disease exacerbated. The similarities and differences between this case and other previously reported cases are discussed.

A 73-year-old woman was admitted to Showa University Fujigaoka Hospital because of massive proteinuria and purpura of the lower extremities. She had received the diagnosis of chronic hepatitis C twenty years earlier. In April 1998, brownish purpura appeared over the lower extremities. A biopsy specimen of the skin showed leukocytoclastic vasculitis. In May 1998, she received the diagnosis of cryoglobulinemia due to the presence of serum cryoglobulin. Purpura improved rapidly when prednisolone (PSL, 30 mg/day) was begun. PSL was discontinued after 54 days of treatment but exacerbation of purpura plus massive proteinuria prompted her first admission to this hospital 36 days later. Physical examination on admission revealed an acutely ill old woman with hypoxia and anemia. Arterial blood gas showed PaO₂=60.7 mmHg. A moderate decrease in her renal function was noted: urinary protein excretion was 2.8 g/day and the serum creatinine level was 1.5 mg/dl. The cryocrit was as high as 60%. Percutaneous renal biopsy was performed ten days after admission. Light microscopy showed glomerular lobulation, hypercellularity in the mesangium and double contour of the glomerular basement membranes (GBM) (Fig. 1A). Immunofluorescence microscopy showed staining for C3, C4, and C1q along the GBM and the mesangium (data not shown). Electron microscopy revealed subendothelial electron-dense osmiophilic deposits and mesangial interposition (Fig. 1B and 1C). Taken together, these membranoproliferative glomerulonephritis (MPGN) features are typical of cryoglobulinemic glomerulonephritis (4). Four sessions of cryofiltration were carried out. As a result, the cryocrit fell to 20%. Her proteinuria was resolved partially. From September 3, the patient was given 1.0 g methyl-prednisolone per day for three successive days, followed by 50 mg/day prednisolone. Her generalized fatigue and purpura improved. On admission, her HCV RNA concentration was 19 kIU/ml but this rose to >850 kIU/ml by September 25. She was then given 9 million units INFα per week to control the replication of hepatitis C virus. HCV RNA fell to 12 kIU/ml on October 22. At discharge, the cryocrit was 12% and partial pressure of oxygen breathing room air was 91.3 mmHg. The dose of PSL was 25 mg/day. INFα was discontinued in January 2002 because of generalized fatigue and numbness of the tongue. Tapered PSL was discontinued on March 1, 2002, when the HCV RNA was 850 kIU/ml.

On March 8, 2002, she was readmitted to the hospital because of exacerbation of purpura, appetite loss and shortness of breath on effort. She was alert on admission, her body weight was 57.0 kg but her height could not be determined. Her blood pressure, pulse rate, and body temperature were 162/82 mmHg, 103 bpm and 36.9ºC, respectively, and the respiratory rate was 25/min. Anemia and peripheral edema were noted. She had a diffuse purpura over both extremities proximal to the knees. Laboratory data on admission are summarized in Table 1. In brief, a complete blood cell count, electrolytes and hepatobiliary enzymes were within the normal ranges. CRP was 2.8 mg/dl, CH50 was <12.0 and RF was 6,680.0 IU/ml. The serum IgM level was 1,110 mg/dl. The cryocrit was 40%. In room air, oxygen saturation was 93.7%. Urine examination showed 0.5 g/day proteinuria with +1 occult blood. A chest radiograph showed bilateral pleural effusion.

The initial diagnostic impression was congestive heart failure. Oxygen saturation remained >97.0%, breathing 2 liters oxygen and on furosemide. On the second hospital day, the patient suddenly complained of dyspnea. There was a loss of 2 kg body weight. Oxygen saturation fell to 87.6% on 10 liters oxygen. Mechanical ventilation was begun to maintain acceptable gas exchange. On mechanical ventilation iSIMV, tidal volume 450 ml, rate 20/min and FiO₂ 0.6, oxygen saturation was 98.4%. A chest radiograph revealed diffuse, bilateral alveolar infiltrates (Fig. 3). A Swan-Ganz catheter was inserted to evaluate the hemodynamic status, disclosing PA 26/15 (20) mmHg, RA 5 mmHg, Pcw 8 mmHg. Furthermore, PO₂/FiO₂ was <200 Torr. These data fulfilled the diagnostic criteria of ARDS (5), except for the sudden onset, and ruled out congestive heart failure. An immune-mediated mechanism of cryoglobulinemia was considered the likely cause of ARDS. The patient was given 1.0 g methyl-prednisolone per day for three successive days, one session of body fluid removal by ECUM and two sessions of plasma exchange. Thereafter, furosemide and 40 mg prednisolone were given. Purpura and respiratory distress were resolved in a few days. On the 6th hospital day, mechanical ventilation was discontinued. The cryocrit fell to 2% on the 8th hospital day. INF therapy was begun on the 20th hospital day. Her general status remained stable for more than 7 months.

We presented a rare case of mixed essential cryoglobulinemia with severe pulmonary involvement. Respiratory insufficiency, which occurred one day after the second admission, raised the possibility that our patient had congestive heart failure. Hemodynamic studies by means of a Swan-Ganz catheter denied this possibility. Acute respiratory distress syndrome (ARDS), originally described by Ashbaugh et al (6), needed to be ruled out. According to the American-European Consensus Conference, the diagnostic criteria of ARDS include 1) acute onset, 2) ratio of the partial pressure of arterial oxygen to the fraction of inspired oxygen [PO₂/FiO₂] <200 mmHg, 3) bilateral infiltrates on chest radiography, and 4) a pulmonary wedge pressure <18 mmHg, or absence of clinical evidence for left-sided heart failure (6). The present patient's severe illness fulfilled all of these criteria except for an acute onset since it is fair to think she already exhibited lung involvement of immune-mediated vasculitis during the first admission. Table1 indicates that PaO₂/FIO₂=278.9 on the second admission. Thus, the present case could be interpreted as acute lung injury (ALI) rather than ARDS, although cytological and/or histopathologic examinations by means of lung biopsy and alveolar lavage were not performed. The etiology of ALI in this case, however, remained to be elucidated. Observations, such as CRP value and negative sputum culture, indicated that ALI was not associated with infection. Patients with cryoglobulinemia may have unknown fever in the absence of infection (7). Thus, low-grade fever could be explained by cryoglobulinemia. Probably, ALI was caused by immune-mediated vasculitis following cryoglobulin deposition. Renal biopsy findings, response to methylprednisolone pulse therapy, and the fact that discontinuance of the PSL therapy lead to her first admission indirectly supported this speculation. A case of systemic lupus erythematosis (SLE) and ARDS presumably caused by immune-mediated pulmonary vasculitis is reported by Domingo-Pedrol et al (8).

Previous reports on severe lung involvement of cryoglobulinemia are summarized in Table 2. A case of mixed cryoglobulinemia with extensive lung disease was presented at a weekly clinicopathologic conference held in 1976 in Barnes and Wohl Hospitals (9). Another case of mixed essential cryoglobulinemia was reported in 1981 by Chejfec et al, with respiratory arrest following acute pneumonia (1). Although the concept of ARDS was not established at that time, the rapid clinical course and pathologic findings at autopsy indicated that these two patients might have died of ARDS. Autopsy of these cases disclosed pulmonary perivascular localization of IgG and IgM. In the latter case, materials containing IgM (kappa) and IgG occluded the pulmonary vessels. The first case of mixed cryoglobulinemia and coincident ARDS was reported in 1989 by Stagg et al (10). Renal manifestations were not described in this report. The patient died of chronic intestinal vasculitis, which appeared about four months after the onset of ARDS. A case of mixed cryoglobulinemia due to hepatitis C infection was reported in 1995 by Roithinger et al, in which the patient died of acute respiratory failure in spite of steroid and antibiotic therapy (2). Diffuse pulmonary vasculitis was found at autopsy. A case of mixed cryoglobulinemia due to Sjögren's syndrome, and in which dyspnea, hypoxia and pulmonary embolism were seen, was reported in 1997 by Konishi et al (11). Dyspnea resolved in one week after 60 mg/day PSL administration. They assumed the pulmonary symptoms were caused by vasculitis associated with cryoglobulinemia.

To the best of our knowledge, this is the first report of mixed cryoglobulinemia due to chronic hepatitis C infection with severe pulmonary involvement in which early and long-term management (i.e. a combination of mechanical ventilation and methylprednisolone pulse therapy followed by PSL administration, plasma exchange and ECUM) were successful. These therapeutic approaches were directed not only to optimal management of ALI but also to reduction in serum cryoglobulin. Although methylprednisolone pulse therapy in ARDS remains controversial (12), it is an established therapy in cryoglobulinemia (13). It has been proposed that the therapy increases the efficacy of anti-inflammatory effects with the least side effects (14) and that removal of circulating cryoglobulin prevents its intravascular precipitation and restores the functional capacity of the overloaded reticuloendothelial system (14, 15). PSL administration after the pulse therapy could have prevented a rebound phenomenon after plasma exchange. Furthermore, PSL also could have prevented pulmonary fibrosis after ARDS (12, 16).

It is generally believed that pulmonary edema in ARDS is due to increased vascular permeability, that intravascular hydrostatic forces may be a factor contributing to the disease progression, and that diuresis and fluid restriction is effective during the first few days after onset (12). Thus, ECUM therapy with central hemodynamic monitoring, which we conducted, might be warranted. We believe that these therapeutic approaches could work in concert and contribute to the patient's recovery. Although the present patient has been free of chest symptoms for more than 7 months, a longer follow-up may be necessary to assess the potential danger of relapse.