Vitamin B for treating peripheral neuropathy
Abstract
This is a protocol for a Cochrane Review (Intervention). The objectives are as follows:
Our objective is to review systematically the evidence from randomised controlled trials of vitamin B complex for treating peripheral neuropathy.
Background
The term 'peripheral neuropathy' has been used to cover any disorder of the peripheral nervous system but this review will focus on generalized peripheral neuropathies. These are disorders of the peripheral nerves which may affect the sensory, motor or autonomic functions. Peripheral neuropathy may present as a chronic and disabling condition; in rare cases it can be acute and fatal. It is a common condition seen by general practitioners and specialists alike. The prevalence is estimated at 2400 per 100,000 (2.4%), increasing with age to 8000 per 100,000 (8%) (Hughes 2002). The common causes are diabetes, alcohol, human immunodeficiency virus infection, and, in some parts of the world, leprosy. The two major types of peripheral neuropathy are demyelination and axonopathy, depending on the predominant involvement of the myelin sheath or the axon, respectively (Kraft 1988).
Pharmacological interventions to prevent, delay or reverse the damage to the peripheral nervous system remain vague and non‐standardized. The availability and relative affordability of vitamin B complex makes this drug a frequent choice for treating peripheral neuropathy. Solid evidence in the literature on the effectiveness of vitamin B complex, however, is lacking. The vitamin B complex is a group of water‐soluble compounds that differ in chemical structure and biological action. Traditionally, the vitamin B complex includes: thiamine (vitamin B1), riboflavin (vitamin B2), nicotinic acid (vitamin B3), pyridoxine (vitamin B6), biotin, pantothenic acid (vitamin B5), folic acid (vitamin B9), cyanocobalamin (vitamin B12), para‐aminobenzoic acid, inositol, and choline (Marcus 1996).
The vitamin B complex functions as coenzymes in several intermediary metabolic pathways for energy generation and blood cell formation which cannot be explained simply. Thiamine is converted to thiamine pyrophosphate (TPP) that functions in carbohydrate metabolism as a coenzyme in the decarboxylation of alpha‐keto acids and alpha‐ketoglutarate and in the utilization of pentose in the hexose monophosphate shunt. Thiamine pyrophosphate appears to play a role in the transmission of nerve impulses. Riboflavin is converted into flavin mononucleotide (FMN) and flavin adenine dinucleotide (FAD) that serve as coenzymes for respiratory flavoproteins. Nicotinamide adenine dinucleotide (NAD) and nicotinamide adenine dinucleotide phosphate (NADP), the active forms of nicotinic acid, are coenzymes for proteins that catalyze oxidation‐reduction reactions in tissue respiration. Pyridoxine, pyridoxamine, and pyridoxal are converted to pyridoxal phosphate that is involved in the metabolic transformations of amino acids and in the metabolism of sulfur‐containing and hydroxyl‐amino acids. Pyridoxal phosphate is required for the synthesis of serotonin and norepinephrine; it appears to be required for the synthesis of sphingolipids for myelin formation as well. Cobalamin is converted to methylcobalamin and 5‐deoxyadenosylcobalamin which are vital in cell growth and replication. Folic acid is converted to several coenzymes essential in cellular metabolism, including the synthesis of some deoxyribose nucleic acid components and normal red blood cell formation.
Because of the vital role of vitamin B complex in energy metabolism, deficient states are reflected in rapidly growing tissues including the nervous tissue due to its high‐energy demand or specific effects of the vitamin (Chaney 1992). Common deficiency symptoms include peripheral neuropathy, depression, mental confusion, lack of motor co‐ordination, and malaise. Specific vitamin B deficiency diseases in humans include beriberi (thiamine), pellagra (nicotinamide), megaloblastic anaemia (folic acid), and pernicious anaemia (cobalamin). The required daily intake of vitamin B for a normal adult is as follows: thiamine 1.0 to 1.5 mg, riboflavin 1.2 to 1.7 mg, pyridoxine 1.4 to 2.0 mg, and niacin 13 to 19 niacin equivalents (Chaney 1992); cobalamin 3 to 5 mg and folic acid 50 mg (Hillman 1996). The therapeutic dose of vitamin B is higher for the treatment of certain deficiency states with neuropathy, such as: 40 mg oral thiamine per day for thiamine deficiency, 10 to 20 mg pyridoxine per day for peripheral neuritis induced by isoniazid (Olson 1996).
While the indications of vitamin B complex for certain conditions such as thiamine for the treatment of Wernicke's encephalopathy (RodriguezMartin 2002), and folic acid for the treatment of anaemia (Mahomed 2002) and the prevention of neural tube defects (Lumley 2001) are well established, their role in treatment of peripheral neuropathies has not been well documented. Methylcobalamin seems to have a stimulating effect on RNA synthesis in neurons and protein synthesis in Schwann cells through a transmethylation process; another possible mechanism of action is through an increase in phosphatidylcholine synthesis in neurons (Goto 1981). In another study, methylcobalamin appears to promote axonal transport and phospholipid synthesis (Devathasan 1986).
Outcome measures in peripheral neuropathy may be measured at the level of pathology (including neurophysiology), impairment, disability and handicap. Impairment may be measured by assessing single items, such as walking speed, or time to place and replacing the pegs in a nine‐hole pegboard, or as a score summing the whole neurological examination, such as the neuropathy impairment score (Dyck 2005). Symptoms may be assessed with a symptom score, and in particular pain may be assessed with a visual analogue scale (Huskisson 1974). In painful neuropathy, assessment of pain is intuitively the most important outcome measure. In other neuropathies, most trials have assessed impairment rather than other measures. Green et al. emphasized the role of electrophysiological tests in determining outcomes in clinical trials involving patients with diabetic peripheral neuropathy (Greene 1981). It is also possible to use quality of life measures such as the Medical Outcomes Study Short Form 36, the Nottingham Health Profile, the Sickness Impact Profile, and the World Health Organization Quality of Life instrument (Abresch 2001).
Objectives
Our objective is to review systematically the evidence from randomised controlled trials of vitamin B complex for treating peripheral neuropathy.
Methods
Criteria for considering studies for this review
Types of studies
We will include randomised and quasi‐randomised studies in which vitamin B is compared with placebo, another treatment or no treatment.
Types of participants
We will include all participants with generalized peripheral neuropathy diagnosed on the basis of symptoms and neurological impairment or symptoms and abnormal neurophysiological results. Diagnosis based on symptoms alone, impairments alone or abnormal neurophysiological results alone will not be sufficient. Symptoms include subjective complaints of pain, tingling, numbness, weakness, etc. Impairments, on the other hand, are measured on neurological examination by testing sensory loss, weakness, etc. or detected on nerve conduction studies.
Types of interventions
We will include vitamin B complex containing thiamine, riboflavin, nicotinic acid, pyridoxine, or cobalamin (or their derivatives, singly or in combination) given in any dose, by any route, for at least 2 weeks. We will not include folic acid unless it is a part of a combined preparation containing other vitamin B compounds.
Types of outcome measures
Changes and number of events occurring over time will be re‐scaled to 'change or number of events in a fixed period' e.g. 3 months, to permit pooling of studies with differing follow‐up periods.
Primary outcome
The primary outcome measure will be short‐term (three months or less) change. For painful neuropathy the primary outcome will be change in pain intensity measured with a validated scale such as the Visual Analog Scale (Huskisson 1974). For non‐painful neuropathy the primary outcome will be change in impairment measured by a validated scale such as the Neuropathy Impairment Score (Dyck 2005).
Studies of these two types of neuropathies will be analysed separately and only combined if an appropriate standardisation can be devised to make the two outcomes equivalent.
Secondary outcomes
(1) Long‐term (after more than three months) change in pain intensity or impairment measured as for the primary outcome.
(2) Short‐term and long‐term (defined as above) change in neuropathic symptoms measured by a validated scale (Dyck 2005).
(3) Short‐term and long‐term (defined as above) change in nerve conduction parameters using the:
(a) amplitude of sensory nerve action potential of the median and sural nerves;
(b) motor conduction velocity of the common peroneal nerve. If data for the common peroneal nerve are not provided then data from other motor nerves will be used.
(4) Serious adverse events as a result of treatment within three months and after three months. Serious adverse events are those which are life threatening, prolong or require hospitalisation, or lead to death.
Search methods for identification of studies
We will search the Cochrane Neuromuscular Disease Group Register for randomised trials using the following search terms: 'Vitamin B' or 'Vitamin B complex' and its different forms, 'aminobenzoic acid', 'biotin', 'folic acid', 'tetrahydrofolate', 'inositol', 'nicotinic acid', 'niacin', 'niacinamide', 'pantothenic acid', 'riboflavin', 'thiamine', 'cobamide', 'cobalamin', 'pyridoxine', and 'peripheral nervous system diseases' and its synonyms, 'peripheral nervous system disorders', 'neuropathies', 'peripheral neuropathies', 'neuritis', 'neuralgia', 'polyneuropathies', 'generalized peripheral neuropathies'. We will also search MEDLINE (from January 2000 to the present) using the strategy given in Table 1, and we will adapt this strategy to search EMBASE (from January 1980 to the present) and local databases: namely, Herdin (Copyright 1997), Philippine Index Medicus (1987 to the present ), ULP, FilDOC, SEAMIC.
| OVID MEDLINE |
| OVID MEDLINE |
Other search strategies
We will review the bibliographies of the randomised trials, quasi‐randomised trials, and clinical trials identified, contact the authors and known experts in the field and approach pharmaceutical companies to identify additional published or unpublished data.
Data collection and analysis
Two authors will independently review the titles and abstracts of all the articles identified by the literature search. The articles will be reviewed for relevance to the research problem and eligibility based on specified inclusion criteria. The authors will discuss the abstracts to resolve any disagreement regarding relevance. A third author will be invited to arbitrate if there is failure to resolve the disagreement. The full text of studies, agreed upon by the authors to be relevant, will be retrieved from the full list of titles and abstracts. All non‐English articles will be translated into English prior to detailed review.
All full text articles will be independently reviewed by the authors to see if they meet the inclusion criteria. We will contact the authors of trials for missing information.
Quality review
The authors will independently appraise the quality of the articles using a 4‐item quality assessment instrument that evaluates the following areas: secure method of randomisation; concealment of allocation; blinding; and completeness of follow‐up (Annane 2004). After independent evaluation, the authors will convene and discuss individual ratings based on the quality scale. If there are inter‐rater discrepancies on the quality evaluation, these will be documented and the authors will further discuss the issue until it is resolved. When the authors cannot reach an agreement, an independent author will be called in.
Data extraction
Data in the articles that meet the inclusion criteria and quality standard (agreed by author consensus) will be extracted independently by two authors. Data for extraction will include study name, design, sample size, study duration (including follow‐up period), participant characteristics (age, sex), inclusion and exclusion criteria, intervention including dosage, route, and treatment duration, outcomes and number of patient drop‐outs, withdrawals, and number of patients analysed in the different treatment groups (Clarke 2001). Data will be extracted onto a specially designed data extraction form. The authors will meet to discuss the content of the articles. Discrepancies will be settled by further discussion until agreement is reached. If disagreement persists, an independent author will be asked to help reach agreement.
Statistical methods
Quantitative analyses of outcomes will be based on intention‐to‐treat results. For continuous outcomes, mean differences in the outcome measures with their 95% confidence intervals (CIs) will be used to quantify the effects of vitamin B (change in pain intensity, change in frequency or severity of neuropathic symptoms and impairments, changes in nerve conduction parameters). The mean difference is the difference in the magnitude of effect in the treatment and control (placebo) groups. To combine and summarize these measures across the studies, weighted mean difference for each outcome measure will be calculated including the 95% CI. For dichotomous outcomes, relative risks (RRs) with 95% CIs will be calculated for individual studies. Trials with different types of outcomes such as different scales measuring pain severity will be combined by determining the difference between the average changes or by converting ordinal outcomes to dichotomous outcomes.
To assess the amount of heterogeneity among the studies, the I squared statistic will be computed in Review Manager 4.2 (RevMan). An I squared value greater than 50% will be taken to mean that a significant amount of heterogeneity exists among the studies. If the studies are found to be generally heterogeneous, the assumptions of the fixed effect model underlying the calculation of the weighted mean difference may not be satisfied. Therefore, analysis based on the random effects model will be selected instead, as this is more conservative and appropriate. Heterogeneity due to one or two unusual and unrepresentative studies will be dealt with by omitting those studies from the calculation of the pooled estimate.
Sensitivity analysis will be undertaken with and without trials lacking high ratings for individual quality attributes.
We will consider separate subgroups of painful and painless neuropathy and of diabetic, alcoholic and other causes of neuropathy. Patients with a documented particular vitamin B deficiency will be considered in a separate subgroup. We will do a subgroup analysis on trials where vitamin B was compared with active treatments and where vitamin B was compared with negative controls with placebo or no treatment. In our discussion we will consider the adverse events reported in the trials identified in the context of non‐randomised evidence from sources such as Meyler's Side Effects of Drugs (Dukes 2000). We will also consider the cost‐effectiveness of treatment, drawing on non‐randomised evidence when necessary.
| OVID MEDLINE |
| OVID MEDLINE |
