Prospective study of guideline-tailored therapy with direct-acting antivirals for hepatitis C virus-associated mixed cryoglobulinemia

PATIENTS

In this prospective study, 44 consecutive patients were enrolled at the outpatient clinic of the Interdepartmental Center for Systemic Manifestations of Hepatitis Viruses, University of Florence, and at the Referral Center for Mixed Cryoglobulinemia, Sapienza University of Rome. Inclusion criteria were detectable levels of serum HCV RNA, presence of an active cryoglobulinemic vasculitis diagnosed as described,35 and eligibility for treatment with a DAA according to the Italian Medicines Agency's indications: “severe HCV-associated extrahepatic manifestations (cryoglobulinemic syndrome with organ damage and B-cell lymphoproliferative syndrome).” This study was conducted according to the ethical guidelines of the 1975 Declaration of Helsinki; treatments were administered on-label and therefore did not require ethical approval. Laboratory studies beyond routine clinical care were approved by the local institutional ethics committees; informed consent was obtained from all patients.

TREATMENT

Patients were treated with different sofosbuvir-based DAA combinations (Table 1) individualized according to the recommendations of the European Association for the Study of the Liver36 on which the Italian Association for the Study of the Liver and the Italian Medicines Agency guidelines are based. Drug dosages were sofosbuvir 400 mg daily, simeprevir 150 mg daily, ledipasvir 90 mg daily, daclatasvir 30 mg daily, and ribavirin 800-1200 mg daily in a weight-based dose. Details on treatment schedules are provided in Supporting Table S1. Two patients with severe vasculitis received rituximab at a reduced dosage of 250 mg/m² given twice 1 week apart37 concurrently with DAA therapy. Patients were followed up for a median of 42 weeks (range 27-53) after the completion of therapy.

CLINICAL AND LABORATORY ASSESSMENTS

HCV infection was proven by detecting circulating anti-HCV antibodies (EIA-2 and RIBA-2; Ortho Diagnostic Systems, Raritan, NJ) and HCV RNA (AMPLICOR HCV Test, v2.0; Roche Diagnostics, Alameda, CA). The HCV genotype was assessed by the VERSANT HCV Genotype 2.0 assay (Siemens Healthcare Diagnostics, Deerfield, IL). SVR was defined as undetectable serum HCV RNA at weeks 12 (SVR12) and 24 (SVR24) after the end of treatment.

Liver disease was evaluated by noninvasive methods including liver elastography using FibroScan (Echosens, Paris, France), imaging, clinical presentation, and laboratory data; the METAVIR score as well as the Model for End-Stage Liver Disease scoring system were used to assess the severity of liver disease.

Clinical response of cryoglobulinemic vasculitis was evaluated by two different methods. One method was a modified version of a classification of clinical response of cryoglobulinemic vasculitis.17, 38-40 Thus, we classified as “full-complete responders” (criterion 1) patients with disappearance of all the baseline symptoms, as “complete responders” (criterion 2) those with improvement of all the baseline symptoms, as “partial responders” (criterion 3) those with disappearance or improvement of at least half of the baseline symptoms, and as “nonresponders” (criterion 4) patients with disappearance or improvement of fewer than half of the baseline symptoms. Clinical response of cryoglobulinemic vasculitis was also evaluated by the Birmingham Vasculitis Activity Score (BVAS), version 3.41

Immunological parameters for evaluating MC response included quantification of cryoglobulins (expressed as percent cryocrit) and measurement of the C3 and C4 complement fractions and of rheumatoid factor. The immunological response was defined as “complete” in case of disappearance of cryoglobulins, “partial” for a decrease of cryocrit to <50% of baseline level, and “null” in all other cases.17, 37, 40 Monoclonal B-cell lymphocytosis was evaluated by flow cytometry, as described.13

All patients were evaluated for clinical and laboratory parameters monthly during the first 2 months of treatment, at the end of treatment (EOT), and at weeks 12 and 24 after the end of treatment.

The molecular analysis of B-cell receptors is described in Supporting Information.

STATISTICAL METHODS

Statistics were performed using the Student t test and Wilcoxon test for quantitative data, chi-squared test, Fisher's exact test for qualitative variables, and Spearman's rank test for correlation between variables. Categorical variables are expressed as count and percentages and quantitative ones as mean ± standard deviation. All statistical tests were two-tailed with a significance level of 0.05 (Stata, version 9.0, and SPSS, version 17.0).

HEPATOVIROLOGICAL RESPONSE

In all patients HCV RNA was undetectable at week 4 of treatment and remained negative throughout the follow-up, with SVR12 and SVR24 rates of 100%. An overall improvement was observed for the main liver function parameters: the mean alanine aminotransferase level decreased from 77.7 ± 77.2 IU/L at baseline to 27.4 ± 10.3 IU/L at SVR12 (P < 0.0001) and to 27.3 ± 10.3 IU/L at SVR24 (P < 0.0001). The mean aspartate aminotransferase decreased from 55.2 ± 60.4 IU/L to 21.3 ± 9.1 IU/L (P < 0.0001) at SVR12 and to 22.6 ± 8.3 IU/L (P < 0.001) at SVR24. A slight decrease of the Model for End-Stage Liver Disease score was observed in the 17 patients with F4 fibrosis from a mean of 9.9 ± 2.5 at baseline to 8.8 ± 2.7 at SVR12 and to 7.3 ± 1.9 at SVR24.

CLINICAL EFFICACY

Among the 44 patients, at SVR12, 34% were full-complete responders and cleared all manifestations of vasculitis (criterion 1); 32% were complete responders, with improvement of all the manifestations of vasculitis (criterion 2); 27% were partial responders, with disappearance or improvement of at least half of the manifestations of vasculitis (criterion 3); and 7% were nonresponders, with disappearance or improvement of fewer than half of the baseline symptoms (criterion 4) (Fig. 1A). At SVR24, all 44 patients had a clinical response of vasculitis, with 36% full-complete responders, 41% complete responders, and 23% partial responders (Fig. 1A). In no patients did the vasculitis manifestations maintain the same pretreatment severity or worsen. The mean BVAS decreased from 5.41 (±3.53) at baseline to 2.35 (±2.25) (P < 0.001) at week 4 on treatment, to 1.39 (±1.48) (P < 0.001) at SVR12, and to 1.27 (±1.68) (P < 0.001) at SVR24 (Fig. 1B). Because the presence of type 2 cryoglobulins may qualitatively and quantitatively correlate with MC vasculitis,42 we compared clinical presentation and response to therapy in patients with type 2 (n = 29) or type 3 (n = 15) in our series. We could not detect any significant difference between the two groups either in the prevalence of specific clinical manifestations at baseline (data not shown) or in the clinical response to therapy; in fact, at SVR24 of the 29 patients with type 2 MC, 35% fulfilled criterion 1, 41% criterion 2, and 24% criterion 3, while these figures were 40%, 40%, and 20%, respectively, among the 15 patients with type 3 MC. Also, we did not notice any trend to a different outcome in patients previously treated with IFN-based therapy, infected with specific HCV genotypes or receiving specific DAA combinations (data not shown).

The relative increase of complete versus partial responders and the disappearance of nonresponders at SVR24 compared to SVR12 (Fig. 1A) suggest that some manifestations of MC vasculitis slowly respond to eradication of HCV. Indeed, the improvement of fatigue, sicca syndrome, and peripheral neuropathy was relatively slow, whereas other manifestations such as palpable purpura, kidney disease, and skin ulcers tended to respond more rapidly (Table 2). Six patients presented with multiple skin ulcers; in 4 of them the ulcers were chronic and resistant to intensive medication and antibiotics, whereas in 2 other cases they occurred early during treatment. In all of these patients ulcers disappeared by SVR24.

Among the 4 patients with renal involvement, the 1 with membranoproliferative glomerulonephritis experienced an increase in the glomerular filtration rate up to normal values, subsiding of proteinuria, and normalization of blood pressure; another patient had a complete remission of the nephrotic syndrome with no further requirement for albumin and diuretics, and the remaining two patients showed a marked improvement of the glomerular filtration rate with normalization of proteinuria.

Two patients, 1 with chronic skin ulcers and 1 with kidney disease, were treated with rituximab, 250 mg/m² given twice 1 week apart, in association with a DAA. In the first patient ulcers completely healed by week 8, and in the other patient complete remission of the nephrotic syndrome was observed at week 4 of treatment. In both patients the remission of vasculitis persisted throughout a 24-week follow-up after EOT.

To clear our data of the confounding effects of rituximab cotreatment, we separately evaluated the 42 patients who received DAA only. The results of this restricted analysis did not substantially differ from those of the analysis performed on the entire cohort, with 33% full-complete responses, 43% complete responses, and 24% partial responses at SVR24 and a decrease of cryocrit from 7.1 ± 15.8% to 3.0 ± 7.5% at SVR12 and to 1.9 ± 5.2% at SVR24 (P < 0.001).

Interestingly, in two patients with MC and lymphoma there was a dissociation of the hematologic response and the response of MC vasculitis. In fact, while in both cases there was a partial regression of the symptoms of vasculitis and a decrease of cryocrit at SVR24, in neither case was there a reduction of monoclonal B-cell lymphocytosis. Notably, in 1 of these patients, who had a CLL-like disorder, treatment with PEG-IFN done 18 months earlier had induced a virologic response associated with regression of both MC vasculitis and monoclonal B-cell lymphocytosis; later virologic relapse was associated with relapse of MC vasculitis and of monoclonal lymphocytosis, leading to treatment with a DAA. We tested whether there were differences in the B-cell receptors expressed by circulating monoclonal B cells or if there were oligoclonal rather than monoclonal expansions before treatment with PEG-IFN or with DAA; at both time points there was a single population of monoclonal B cells expressing an idiotype encoded by somatically mutated IGHV3-7 heavy chain and IGKV3-15 light chain variable genes, homologous to the WA cross-idiotype.43, 44 Thus, different responsiveness of monoclonal B-cell lymphocytosis to PEG-IFN compared to DAA therapy appears to be more likely related to the type of treatment than to changes in the characteristics of lymphoma cells. The other patient, who had marginal zone lymphoma with plasmacytic differentiation, had a single circulating monoclonal B-cell population expressing an idiotype encoded by IGHV2-5 and IGKV1-5. Detailed information on these patients is provided in Supporting Information.

IMMUNOLOGICAL EFFICACY

The mean cryocrit level rapidly decreased from 7.2 ± 15.4% at baseline to 2.9 ± 9.0% at week 4 on treatment (P < 0.05) and remained substantially stable up to SVR12 to eventually decrease to 1.8 ± 5.1% (P < 0.001) at SVR24 (Fig. 2A). The immunological response at SVR12 was complete (disappearance of cryoglobulins) in 32% of cases, partial in 39%, and null (<50% decrease of cryocrit) in 29%. At SVR24, cryocrit was available for 38/44 patients and there were 39.5% complete responders, 34% partial responders, and 26.5% null responders; in patients with null response, cryocrit values decreased at least 38% from baseline. The serum C4 levels were serially measured only in the 14 patients out of 24 who had low values at baseline (Fig. 2B); in 45% of these patients C4 levels had returned to normal at SVR12. Five of the 11 patients studied at SVR24 had persistently low C4 levels, and 4 of them had persistent cryoglobulins. Rheumatoid factor levels decreased from 131.2 ± 252.9 at baseline to 66.2 ± 91.8 at SVR12 and to 39.0 ± 29.8 at SVR24 (Fig. 2C).

Reduction of cryocrit, expressed as a percent decrease from baseline, did not correlate with the degree of clinical response, expressed as a percent decrease of BVAS, either at SVR12 (rho = 0.059 by Spearman rank correlation) or at SVR24 (rho = 0.29). Disappearance of cryoglobulins and regression of all manifestations of vasculitis (criterion 1) did not appear to be associated; in fact, at SVR24 there were only 8 negative cryocrits out of the 16 patients fulfilling criterion 1.

TREATMENT SAFETY

Adverse events related to DAA occurred in 26/44 (59%) patients; they were generally mild and in no instance led to discontinuation of DAA treatment. Most patients (22/26) had more than one adverse event, with anemia, fatigue, and nausea being the most complained. Anemia occurred in 13 patients, all receiving ribavirin. In 8 patients with decrease of hemoglobin to <10 g/dL (>8.5 g/dL) reduction of ribavirin by 200 mg resulted in no further decline. In 4 additional patients hemoglobin decreased further despite the reduction of the ribavirin dose; 3 patients were treated with erythropoietin and completed ribavirin treatment at reduced dosage, while in 1 patient ribavirin was withdrawn while continuing sofosbuvir plus simeprevir. In another patient, an abrupt fall of hemoglobin to 7.5 g/dL required blood transfusion; reduction of ribavirin by 200 mg was followed by stable hemoglobin levels. In 2 patients (1 needing reduction and 1 withdrawal of ribavirin) hyperbilirubinemia also occurred. Further details on adverse events are reported in Table 3.