A total of 491 patients were diagnosed with cryoglobulinemia at Peking Union Medical College Hospital between January 2015 and December 2022. A total of 101 (20.1%) patients had type I cryoglobulinemia, 390 (79.4%) patients had mixed cryoglobulinemia, including 91 (18.5%) with type II, and 299 (60.1%) with type III. A total of 243 (49.5%) patients were asymptomatic, and the remaining patients (50.5%) had at least one organ involved. Overall, 10 (2.0%) patients (2 males and 8 females) fulfilled the inclusion criteria accounted for 2.0% of all cryoglobulinemia patients (n = 491). Patients with pulmonary involvement included 3 with type I cryoglobulinemia and 7 with mixed cryoglobulinemia, including 4 patients with type II and 3 patients with type III. For patients with pulmonary involvement, the median age at diagnosis was 53 years (range, 30–73 years), and the median duration from symptom onset to disease diagnosis was 8 months (range, 2–49 months).

We compared the baseline characteristics of type I cryoglobulinemia patients with (n = 3) and without pulmonary involvement (n = 98). We also compared the baseline characteristics of mixed cryoglobulinemia patients with (n = 7) and without pulmonary involvement (n = 383). The demographic and clinical characteristics are presented in Tables 1 and 2. Compared to type I patients without pulmonary involvement, patients with pulmonary involvement tended to have a higher RF titer (7.0 vs. 4.6 IU/ml, P = 0.065). Out of 390 mixed cryoglobulinemia patients, those with pulmonary involvement had higher frequencies of skin involvement (71.4% vs. 29.2%, P = 0.046), renal involvement (57.1% vs. 20.9%, P = 0.041), and peripheral nerve involvement (42.9% vs. 6.8%, P = 0.011) compared to those without pulmonary involvement. Furthermore, mixed cryoglobulinemia patients with pulmonary involvement showed a markedly higher RF titer than those without pulmonary involvement (226.1 vs. 98.0 IU/ml, P = 0.048).

All patients with pulmonary involvement experienced persistent pulmonary symptoms, which included dyspnea (50%), dry cough (30%), chest tightness (30%), and hemoptysis (10%). The other most common clinical manifestations were purpura (50%) and proteinuria (50%), followed by renal function impairment (40%), hematuria (40%), sensory neuropathy (30%), arthralgia (20%), heart failure (20%), and ulcer (10%). The detailed clinical characteristics of patients with pulmonary involvement are shown in Table 3. Six patients had additional organs affected by cryoglobulinemia. Skin (n = 6) was the most commonly involved organ, followed by kidney (n = 5), peripheral nerves (n = 3), joints (n = 2), and heart (n = 2). Furthermore, both patients with cardiac involvement presented with heart failure. Echocardiography or cardiac magnetic resonance revealed decreased ejection fraction, and massive pericardial effusion in these patients. All type I patients were secondary to B-cell non-Hodgkin lymphoma, including DLBCL (n = 2) and SLL/CLL (n = 1). Among the 7 patients with mixed cryoglobulinemia, 4 patients were essential, and the remaining 3 patients were secondary to infections (n = 2), including HCV (n = 1) and EBV (n = 1), and SLE (n = 1).

The median concentration of cryocrit for patients with pulmonary involvement was 100.9 mg/L (range, 24.3–3463 mg/L), all of which was IgM isotype, including 6 with IgMκ, 1 with IgMλ, and 3 with poly IgM. Serum complement levels were evaluable in all patients with pulmonary involvement: seven patients had a decrease in C3 serum levels, and the median concentration was 0.489 g/L (range, 0.283–0.681 g/L), and five patients had a decrease in C4 serum levels, and the median concentration was 0.016 g/L (range, 0.002–0.076 g/L). Out of 9 patients with pulmonary involvement, 6 patients exhibited elevated titers of RF, and the median level of RF was 463.4 IU/ml (range, 41.5-19489.2 IU/ml).

In the 4 patients with renal function impairment, the median serum creatinine level was 176 µmol/L (range, 99–299 µmol/L). Five patients experienced proteinuria, with a median 24-hour urine protein level was 1.47 g/L (0.62–1.84 g/L), while four patients presented with hematuria. Additionally, serum IFE was positive in 3 patients (#2, #6, and #7) of 9 evaluable patients, all of which were IgMκ isotype, and urine IFE was positive in 2 patients (#1, #2) of 5 evaluable patients, including one light chain of κ and one light chain of λ. Furthermore, among 9 patients who underwent SPE, two patients (#6, #7) were positive, with M protein levels of 10.50 g/L and 21.60 g/L, respectively.

The radiological presentation of patients and examples of typical radiologic findings are shown in Table 4; Fig. 1. Both patients (#1, #2) with DAH showed diffuse bilateral GGO and multiple patches, and patient #2 also presented with pleural thickening and pleural effusions (Fig. 2A-C). Among the other 8 patients, 6 patients showed diffuse bilateral patchy GGOs, which were distributed mainly in the bilateral lower lobes of two patients (#5 and #10). Multiple, diffuse, small nodules were observed in four patients, two of whom (#8 and #9) had a large solitary nodule (maximum diameter > 10 mm) (Fig. 1C). Moreover, the CT scans revealed pleural thickening (50%), pleural effusions (30%), reticulations mainly in the bilateral lower lobes (20%), and bilateral interlobular septal thickening (10%).

A-E, CT imaging findings associated with cryoglobulinemia and lung involvement. A, Diffuse, bilateral, patchy GGOs. B, Diffuse alveolar hemorrhage. C, Solitary large nodule in the right upper lobe. D, Diffuse, bilateral, multiple, small nodules. E, Bilateral pleural thickening. F, Bilateral pleural effusions

CT scans during the treatment of patient #2. A-C, Bilateral multiple patchy and diffuse GGOs along with symptoms of acute dyspnea and hemoptysis. D-F, Remarkable improvement of pulmonary infiltrates after high-dose corticosteroids and rituximab for 4 courses. G-I, Stability of symptoms and infiltrates on repeat chest CT 6 months after the episode of hemoptysis

(A)

PFTs were performed in three patients (#6, #7, and #10) (Table 4). The baseline PFTs of patient #6 revealed restrictive ventilatory dysfunction and diffusing-capacity defects. Likewise, patient #7 was subjected to baseline PFTs and demonstrated mild restrictive ventilatory dysfunction. Patient #10 showed mild restrictive ventilatory dysfunction with normal forced expiratory volume in 1 s (FEV1) and diffusing-capacity defects. After treatment with corticosteroids and mycophenolate mofetil (MMF), the predicted diffuse lung capacity for carbon monoxide (DLCO) value of patient #10 improved (from 46.8 to 67.2%) significantly along with normal restrictive ventilatory function.

The treatments and outcomes of the 10 patients are listed in Table 3. One patient (#3) gave up treatment and died from underlying disease progression. Five (50%) patients received a corticosteroid-based regimen as first-line treatment, including corticosteroids alone (n = 3), corticosteroids and cyclophosphamide (CTX) (n = 1), and corticosteroids and MMF (n = 1). Four (40%) patients were treated with a rituximab-based regimen, including rituximab and corticosteriods (n = 1), R-CHOP (rituximab, cyclophosphamide, vincristine, and prednisone) (n = 1), DRC (rituximab, cyclophosphamide and dexamethasone) (n = 1), and antiviral agents and DRC (n = 1).

Of these 9 treated patients, 7 patients achieved clinical remission, consisting of one (#5) with complete remission (CR) and six with partial remission (PR). Furthermore, all 4 patients treated with rituximab-based regimens achieved clinical remission, while both patients (#4, #8) with no remission (NR) received corticosteroid treatment alone. Of the 5 patients evaluable for laboratory response, three patients (#1, #9 and #10) achieved CR, one (#5) achieved PR, and one (#2) showed NR. Moreover, patient #7 was treated with three lines of therapy and has been stable until now, which we previously reported [19].

The median follow-up was 16 months (range, 2–67 months) for patients with pulmonary involvement. Five patients (50%) with pulmonary involvement died during follow-up. The causes of death included underlying disease progression (n = 3) and therapy-related infection (n = 2). The estimated three-year OS and PFS for all patients with cryoglobulinemia were 94.8% and 84.5%, respectively (Fig. 3A). Patients with pulmonary involvement had significantly worse OS and PFS than those without pulmonary involvement (two-year OS 40% vs. 95.6%, P < 0.0001; two-year PFS 30% vs. 87.7%, P < 0.0001; Fig. 3B, C).

Comparison of survival among patients of cryoglobulinemia. (A) Overall survival (OS) and progression-free survival (PFS) of the whole cohort (n = 491). (B) OS of cryoglobulinemia patients with pulmonary involvement (n = 10) and those without pulmonary involvement (n = 481). (C) PFS of cryoglobulinemia patients with pulmonary involvement (n = 10) and those without pulmonary involvement (n = 481)