Hepatitis C virus (HCV) is a global concern as an estimated 2.2% to 3% of the worldwide population, or 130 to 170 million people, are presently infected (Lavanchy 2009). HCV seroprevalence varies dramatically by geographic location from approximately 25% in Egypt (Arthur 1997), to only 0.01% to 0.1% in the United Kingdom and Scandinavia (Alter 2007). Risk of acquiring HCV is increased by blood transfusion, injection drug use, contaminated medical equipment, birth to an infected mother, and multiple sexual partners (Alter 2007). Chronic infection develops in approximately 70% of HCV-infected people leading to an increased risk of cirrhosis and hepatocellular carcinoma (Global Burden of Hepatitis C Working Group 2004; Lavanchy 2009). Extrahepatic manifestations afflict 40% to 74% of individuals with HCV infection (Acharya 2008). The most common neurologic complication of HCV is peripheral neuropathy (Acharya 2008).
Prevalence of peripheral neuropathy among people with HCV varies depending on the study population and the definition of neuropathy (Cacoub 2005). Standard nerve conduction studies detect peripheral neuropathy in 15.3% of people with HCV; however, approximately one-third have subclinical disease (Santoro 2006). Among people with HCV, prevalence of clinically symptomatic sensory or motor peripheral neuropathy is approximately 10% (Cacoub 2000; Santoro 2006). Small fibre neuropathy detected by pain-related evoked potentials has a much higher prevalence at 43.5% of the HCV-infected population (Yoon 2011). Prevalence of peripheral neuropathy among people with HCV and mixed cryoglobulinemia rises to close to 90% if neuropathy is defined only by the presence of paresthesia (Ferri 1992). Prevalence of mixed cryoglobulinemia among people with HCV is 44% (Kayali 2002). Among people with HCV and neuropathy, 78% demonstrate mixed cryoglobulinemia (Nemni 2003) but the presence of cryoglobulinemia does not increase overall risk of developing a neuropathy (Santoro 2006).
Among people with HCV who have peripheral neuropathy demonstrated using standard nerve conduction studies, approximately half (52.8%) exhibit a symmetrical axonal sensorimotor neuropathy while the remainder have a mononeuropathy or multiple mononeuropathies (Santoro 2006). Among the HCV-infected population affected by neuropathy, prevalence of a symmetrical sensorimotor polyneuropathy with predominant sensory features was significantly higher among those with mixed cryoglobulinemia compared to those without cryoglobulinemia, while among cryoglobulin-negative people with HCV there was a significantly higher prevalence of a well-defined mononeuritic process (Nemni 2003). HCV infection is also associated with polyarteritis nodosa (PAN)-type systemic vasculitis (Cacoub 2001; Kanda 2003). People with PAN-type vasculitis demonstrate a sudden onset sensorimotor multifocal mononeuropathy involving all limbs, exhibiting a severe motor deficit and painful sensory symptoms (Cacoub 2001). In contrast, neuropathy secondary to mixed cryoglobulinemia is a distal sensorimotor polyneuropathy with primarily sensory features. PAN-type vasculitis differs from mixed cryoglobulinemic vasculitis as it affects larger vessels, the infiltrates include neutrophils and it is associated with necrotizing angiitis (Cacoub 2001). Differentiating PAN-type vasculitis from mixed cryoglobulinemic vasculitis is challenging as 90% of people with confirmed PAN-type vasculitis have mixed cryoglobulinemia (Cacoub 2001).
The pathophysiology of neuropathy associated with HCV is not definitively known (Cacoub 2005). However, there are several proposed mechanisms of neuropathy in HCV, including (1) cryoglobulin deposition in the vasa nervorum, (2) HCV-mediated vasculitis resulting in ischemia, and (3) immune-mediated inflammatory demyelination (Stübgen 2009). We will study only the first two categories in this review.
Cryoglobulins are serum immunoglobulins that precipitate when cooled to less than 37°C, dissolving upon reheating (Shihabi 2006). There are three types of cryoglobulins. Type I cryoglobulins are often associated with hematologic malignancy and are composed of a monoclonal immunoglobulin (Ig), while types II and III ('mixed cryoglobulins') are composed of polyclonal IgG associated with either monoclonal (type II) or polyclonal (type III) IgM possessing rheumatoid factor activity.
Cryoglobulinemic neuropathy may occur by small vessel obstruction secondary to cryoglobulin deposition or immune complex-induced vasculitis of the vasa nervorum (Chad 1982). Cryoglobulin deposition within the lumen or vessel walls of the vasa nervorum or both impedes circulation leading to nerve ischemia (Vallat 1980; Chad 1982; Nemni 1988). In addition, endoneurial vessels demonstrate endothelial cell hypertrophy contributing to obstruction of the vessel lumen (Nemni 1988; Bonetti 1997). In mixed cryoglobulinemia, small and medium sized arteries demonstrate perivascular inflammatory infiltrates (Chad 1982). Vasculitis frequently affects arteries supplying the epineurium while involvement of endoneurial vessels is less frequently seen (Nemni 1988). Vasculitic infiltrates of the vasa nervorum are primarily lymphocytic (Cream 1974; Nemni 1988). Infiltrates around vessels supplying the endoneurium include T-lymphocytes and macrophages, while epineurial vessel inflammatory infiltrates consist of monocytes accompanied by T- and B-lymphocytes (Bonetti 1997). Overall there is axonal degeneration which occurs in the absence of primary demyelination (Bonetti 1997). Axonal degeneration may be secondary to ischemia resulting from microcirculation occlusion or vessel destruction resulting from vasculitis (Chad 1982; Nemni 1988; Bonetti 1997).
Detection of HCV RNA in epineurial cells is evidence of viral invasion of peripheral nervous system tissue (Bonetti 1999). The presence of β2-integrin-positive lymphocytes around epineurial vessels, in the absence of immune complex, may suggest a T-cell mediated cause of neuropathy (Bonetti 1997).
The optimal treatment for HCV-related peripheral neuropathy has not been established. Based on the proposed mechanisms of neuropathy in HCV, there could be an effect from antiviral therapy or immune modulation therapy. A consensus committee on the treatment of HCV-related mixed cryoglobulinemia syndrome recommended the use of rituximab, glucocorticoids, apheresis and cyclophosphamide as possible treatments for people with serious complications, such as neuropathy (Pietrogrande 2011). The committee recommended caution in the use of antiviral therapy, as this may be associated with worsening of neuropathy.
Why it is important to do this review
HCV-related neuropathies produce significant morbidity and at present there is lack of clarity regarding the best management of people with this extrahepatic complication of HCV. This review will allow us to analyze the currently available evidence for treatment of cryoglobulinemic or non-cryoglobulinemic peripheral neuropathy associated with HCV infection. People with HCV infection with or without cryoglobulinemia can have a symmetrical sensorimotor polyneuropathy, mononeuropathy or multiple mononeuropathies. These three entities will all be captured under the term 'peripheral neuropathy'. Given the apparent heterogeneity of pathogenic mechanisms and clinical presentation in neuropathy associated with HCV, we will evaluate treatment trials for cryoglobulinemic and non-cryoglobulinemic peripheral neuropathy associated with HCV infection separately during subgroup analysis.
To assess the effects of interventions (including interferon α, interferon α plus ribavirin, corticosteroids, cyclophosphamide, plasma exchange, and rituximab) for cryoglobulinemic or non-cryoglobulinemic peripheral neuropathy associated with HCV infection.
Criteria for considering studies for this review
Types of studies
We will consider randomized controlled trials (RCTs) or quasi-RCTs examining the effects of an intervention versus placebo or another intervention ('head to head' comparison study design) for cryoglobulinemic or non-cryoglobulinemic peripheral neuropathy associated with HCV infection.
Types of participants
We will consider trials including participants of any age with HCV documented using serology, enzyme immunoassay (EIA), recombinant immunoblot assay (RIBA), and/or nucleic acid amplification technology (NAT). Participants must be tested for cryoglobulins, using any validated method, as cryoglobulinemic and non-cryoglobulinemic peripheral neuropathy will be analyzed separately during subgroup analysis. Participants must have a diagnosis of polyneuropathy, mononeuropathy, or multiple mononeuropathy. Neuropathy can be diagnosed based on symptoms, signs, and/or laboratory investigation. We will exclude participants with HCV-associated chronic inflammatory demyelinating polyneuropathy (CIDP). Clinical and laboratory investigations should be sufficient to reasonably exclude other causes of neuropathy.
Types of interventions
We will consider trials of any intervention (including interferon α, interferon α plus ribavirin, corticosteroids, cyclophosphamide, plasma exchange, and rituximab) alone or in combination versus placebo or another intervention ('head to head' comparison study design) evaluated after a minimum interval of follow-up of at least six months. The duration of intervention is anticipated to vary considerably according to specific treatment. We will not specify a minimum duration of intervention, however, reasonable duration of treatment will be determined by author consensus. Treatment for HCV-related peripheral neuropathy prior to enrollment in a given trial will be disclosed where this information is available. We will not consider trials of interventions solely used to treat pain.
Types of outcome measures
We will not use our prespecified outcomes as criteria for selecting studies. If a trial reports an outcome using more than one of the specified scales, we will use the trial's primary outcome measure for our analysis. If the reported scales address one of the trial's secondary outcomes or more than one scale is used for the trial's primary outcome, we will determine which scale to use based on author consensus.
Neuropathy associated with HCV most often has primarily sensory features (Yoon 2011). Therefore, the primary outcome will be the change in sensory impairment at the end of the follow-up period, which will be at least six months following treatment initiation. Sensory impairment can be evaluated using any validated sensory neuropathy scale (including Sensory Sum Score (Merkies 2000) and Utah Early Neuropathy Scale (Singleton 2008)) or quantitative sensory testing.
Secondary outcomes, assessed at the end of the follow-up period, which will be at least six months following treatment initiation, will include.
- Change in impairment, characterized by any validated combined sensory/motor neuropathy scale including the Neuropathy Symptom Score (Dyck 1987), Michigan Diabetic Neuropathy Score (Feldman 1994), Michigan Neuropathy Screening Instrument (Feldman 1994), Neuropathy Impairment Score in the Lower Limbs (Bril 1999), Modified Toronto Clinical Neuropathy Score (Bril 2009), and Total Neuropathy Score (Cornblath 1999).
- Change in disability, assessed using any validated disability scale including Overall Disability Sum Score (Merkies 2002), Neuropathy Disability Score (Young 1993), and Modified Rankin Scale (van Swieten 1988). Although disability is more relevant to patient outcome than impairment, we chose to include disability as a secondary outcome due to perceived paucity of this outcome as a measure in clinical trials concerning HCV-related neuropathy and lack of sensitivity in sensory predominant neuropathy.
- Change in peroneal or tibial compound muscle action potential (CMAP) amplitude, peroneal or tibial motor conduction velocity, or sural sensory nerve action potential (SNAP) amplitude by standard nerve conduction studies.
- Number of patients with improved or resolved symptoms of neuropathy, as determined by a patient or investigator global impression of change or other applied measure of global assessment.
- Number of participants experiencing one or more severe adverse events within the follow-up period. Severe events will be fatal, life-threatening, require or prolong hospitalization, or cause discontinuation of treatment.
Outcomes for inclusion in 'Summary of findings' tables
If there are sufficient data we will include 'Summary of findings' tables to summarize the effects of interventions for key outcomes including change in sensory impairment, change in combined motor/sensory impairment, change in disability, and adverse events. We will rate the evidence for each outcome according to the GRADE approach as high, moderate, low or very low quality and provide the rationale for these decisions. We will consider downgrading evidence for the following reasons: risk of bias, imprecision, indirectness, unexplained heterogeneity, or publication bias. Reasons for upgrading evidence will be: a large magnitude of effect, if all plausible confounding tends to underestimate an apparent intervention effect, and a dose-response gradient (Atkins 2004).
Search methods for identification of studies
We will search the Cochrane Neuromuscular Disease Group Specialized Register using the terms: 'neuropathy', 'peripheral neuropathy', 'peripheral nervous system diseases', 'polyneuropathy', 'mononeuropathy', 'hepatitis C', 'randomized clinical trial', 'controlled clinical trial' and 'drug therapy'. We will search the Cochrane Central Register of Controlled Trials (CENTRAL) (The Cochrane Library, latest issue), MEDLINE (January 1966 to the present) and EMBASE (January 1980 to the present) using a search strategy including the above terms. We will also search trial registries, including ClinicalTrials.gov and the World Health Organization International Clinical Trials Registry Platform (ICTRP) portal for ongoing trials. The Networked Digital Library of Theses and Dissertations (NDLTD) will be searched for any relevant trial not published in peer reviewed literature.
We will search the National Institute for Health Research DARE (Database of Abstracts and Reviews of Effects) and HTA (Health Technology Assessments) databases to identify reviews and assessments. The NHS Economic Evaluation Database (NHS EED) will be searched for cost information concerning interventions for inclusion in the discussion if information is available.
The search strategy for MEDLINE is in Appendix 1.
Searching other resources
We will check all references in the identified trials and contact trial authors to identify any additional published or unpublished data.
Data collection and analysis
Selection of studies
Two review authors will independently review the title and abstract of all studies identified by the search. The review authors will use a data extraction form to record key inclusion and exclusion details from each study. The two authors will determine whether a study meets the inclusion and exclusion criteria for the review or if there is uncertainty whether the study fulfills criteria based on the information available in the abstract. If there is disagreement that cannot be resolved through discussion between the authors, then the third review author will also review the study and the authors will use consensus to determine whether to assess the study in more detail. The same two authors will assess the full text of all studies selected through this process and will make a final decision regarding inclusion of the study using the same method of managing disagreement through consensus, as described above. We will report the articles that are selected for full text assessment but not for data analysis and provide the reasons for exclusion in the final review.
Data extraction and management
Two review authors will independently and in duplicate extract data for the studies to be included in the analysis onto a specially designed data collection form and they will then compare data for completeness and accuracy. The data extracted will include study identification information; details of the study design; number and types of participants, including details of the diagnosis of HCV-related neuropathy, presence of cryoglobulin, and exclusion of other important causes of neuropathy; details of the intervention (type, dose, duration, and comparison agent); assessment of the risk of bias using The Cochrane Collaboration's 'Risk of bias' tool; data for primary and secondary outcomes; and a description of the main conclusions of the trial. For continuous variables we will extract the number of participants, mean values, and standard deviations of the variable at the end of follow-up for the treated and comparison groups. For the dichotomous variables we will extract the number of participants achieving benefit or an adverse effect for each group and the total number of participants in each group.
Assessment of risk of bias in included studies
The two authors reviewing each included study will independently perform an assessment of risk of bias using the 'Risk of bias' tool included in the Cochrane Handbook for Systemic Reviews of Interventions Version 5.1.0 (updated March 2011) (Higgins 2011). The tool will include an evaluation of selection bias, performance bias, detection bias, attrition bias, reporting bias, and other identified sources of bias. For each domain we will judge the possibility of bias as high, low, or unclear. A summary of study bias will be judged as low, high, or unclear. If the summary of bias is unclear then the review authors will request additional information from the study's principal investigator to permit an assessment of bias. If there is disagreement between the two review authors regarding bias grading, then we will use consensus judgment in the manner described above. We will report elements of potential bias for each included study. We will report institutional research ethics board approval of included trials where this information is available.
Measures of treatment effect
We will evaluate each of the various treatments separately. We anticipate that the primary outcome, change in sensory impairment, and the various secondary outcomes will be measured using predominantly continuous variables, such as assessment scales, sensory test scores, and nerve conduction variables. For continuous variables we will calculate the mean differences (MD) with 95% confidence intervals (CIs) if the measures evaluate similar outcomes. If the outcome data are different, then we will calculate standardized mean differences (SMD) with 95% CIs. For dichotomous variables, such as number improved or resolved at the end of the follow-up period and number with adverse effects, we will calculate a risk ratio (RR) and risk difference (RD).
Unit of analysis issues
We are not anticipating significant unit of analysis issues. If there are repeated measurements of outcomes over time, the review will select the final measure as the one most relevant to the long-term course of the patient. We will summate adverse effects as the total number of participants with adverse effects, regardless of time of occurrence or resolution prior to the completion of the trial. We do not expect there to be cluster-randomized trials for this condition. We may encounter cross-over trial design, but this is potentially inappropriate for this condition due to the progressive nature of neuropathy. If we encounter cluster-randomized or cross-over trials, we will evaluate them individually to determine if they are appropriate for inclusion in the analysis. For cross-over trials we are likely to include only the first randomized arm.
Dealing with missing data
If we encounter missing data, we will contact the principal investigator of the trial to provide the missing data. If we are unable to obtain missing data, then we will use replacement values, specifically the last recorded observation will be used in the final analysis. We will perform a sensitivity analysis of the effect of use of the replacement value and we will discuss the impact of this approach in the review.
Assessment of heterogeneity
The main potential heterogeneity in participants will be the presence or absence of cryoglobulins. Heterogeneity in interventions should be minimized by analyzing the different interventions separately. There could be significant heterogeneity in the outcomes measured due to the duration of intervention or follow-up. We may detect heterogeneity due to methodological quality of the studies. We will test for heterogeneity using the Chi
Assessment of reporting biases
To assess for reporting bias of trials, we will search national and international trial registries for completed and unreported trials for HCV and neuropathy. If we find unreported trials, the review authors will request results from the trial's principal investigator. To ensure optimal completeness of identification of published trials, the search methods will include multiple trial databases as noted above and we will search the references of relevant articles. We will use funnel plots to assess for small study bias (Egger 1997). The interpretation of funnel plots where there are small numbers of trials will be done with great caution.
Meta-analysis for the outcomes will use a fixed-effect model if heterogeneity is low or a random-effects model if we detect significant heterogeneity. Meta-analysis for continuous outcomes will involve an analysis of differences in means, whereas for dichotomous outcomes the analysis will assess RR of the outcome. For the dichotomous outcomes, a summary result for each comparison group for each study will be determined. For instance, for the fourth secondary outcome we will use the total number of participants who are described as improved or resolved as the number benefited for that study. For the fifth secondary outcome we will use the total number of participants with a severe adverse effect regardless of type as the number harmed for that study. Based on this, the RR for number benefited and harmed can be calculated across studies. We will discuss costs of the interventions in the discussion if that information is available.
Subgroup analysis and investigation of heterogeneity
We will perform subgroup analysis on participants with or without associated cryoglobulins. We will perform subgroup analysis on short (six months) or long (greater than six months) follow-up. We will use a random-effects model of analysis if we detect heterogeneity. We will analyze homogenous results using a fixed-effect analysis.
We will perform sensitivity analysis on use of fixed-effect or random-effects models. For continuous variables that use different scales to measure the same outcome, we will perform sensitivity analysis on the choice of SMDs, as opposed to MDs. If we use replacement values for missing data, we will perform a sensitivity analysis on the effect of this method. We will perform sensitivity analysis on other issues identified during performance of the review.
Editorial support from the Cochrane Neuromuscular Disease Group is funded by the MRC Centre for Neuromuscular Diseases.
Appendix 1. MEDLINE search strategy
1 randomized controlled trial.pt. (331851)
2 controlled clinical trial.pt. (84651)
3 randomized.ab. (235294)
4 placebo.ab. (132892)
5 drug therapy.fs. (1550727)
6 randomly.ab. (169507)
7 trial.ab. (243740)
8 groups.ab. (1112501)
9 or/1-8 (2882099)
10 exp animals/ not humans.sh. (3755264)
11 9 not 10 (2447589)
12 exp Peripheral Nervous System Diseases/ (115217)
13 (neuropath$ or polyneuropath$ or mononeuropath$).mp. (92962)
14 12 or 13 (168261)
15 exp Hepatitis C/ (42453)
16 (hepatitis c or hcv or hepacivirus).mp. (57103)
17 11 and 14 and 16 (174)
18 remove duplicates from 17 (171)
Contributions of authors
Declarations of interest
None of the review authors have any conflicts of interest to report.
Sources of support
- Dalhousie University and McGill University, Canada.The authors acknowledge Dalhousie University and McGill University for general support provided to the authors while they conducted this review.
- No sources of support supplied