A joint effort: Differentiating types of arthritis

Arthritic variations

Arthritis affects millions of people worldwide. In the UK, The National Health Service has estimated that about 10 millions people suffer from this disease, while the CDC has put the figure in the USA at about 20 million. The term arthritis actually covers more than 100 related diseases, the most common forms being osteoarthritis (OA), rheumatoid arthritis (RA), ankylosing spondylitis (AS) and gout.

In OA, the cartilage between the joints wastes away to allow painful rubbing together of the bones that should be separated. Sufferers of RA have painful swollen joints resulting from attack by their own immune system. AS mostly affects the joints, muscles and ligaments of the spine resulting in pain and inflammation but its cause is unknown. In gout, the big toe is the normal target but other joints can be affected. It results from the build up of uric acid crystals in and around the joints.

Early intervention is central to preventing permanent joint damage, but clinicians often find it hard to diagnose arthritis early. In recent years, a number of research groups have attempted to find metabolites in the body which signal that arthritis is present, even in the early stages before clear symptoms develop. This approach has the bonus of digging out information on the biochemical basis of the disease, to give better understanding of how it develops.

Now, an international team of scientists from China, Hong Kong and the USA have undertaken a comparative metabolomics study of four types of arthritis to see if there are any similarities or differences in the metabolic signatures. They attacked OA, RA, AS and gout.

Blood testing

Blood serum was prepared from 114 patients in total covering the four diseases, plus 60 healthy controls. The samples were analysed by GC/MS following derivatisation with a silylating agent to ensure that polar metabolites could pass through the column. The individual metabolites were identified from their retention times and by matching the spectra with those from a mass spectral database.

The same samples were also analysed by LC/MS with electrospray ionisation in positive- and negative-ion modes to cover the widest range of metabolites. The spectra from these runs were matched against those of reference standards and the Human Metabolome Database.

This combined approach led to the confident identification of 196 metabolites in total and the profiles from each type of arthritis were compared by statistical analysis to reveal both the common and unique features.

Unique metabolite signatures

Two-dimensional discriminant analyses clearly differentiated the arthritic patients from the controls by the different abundances of a number of metabolites. Regardless of the type of arthritis, the levels of six metabolites were consistently altered, suggesting that a number of common metabolic defects are associated with joint inflammation and lesion.

They were homoserine, glyceraldehyde, lactic acid, dihydroxyfumaric acid and aspartic acid, which were elevated in patients, and 4,8-dimethylnonanoylcarnitine which was depleted. Together, they predicted the presence of arthritis with a sensitivity of 81% and a specificity of 88%.

When a biochemical pathway analysis was carried out for this subset, the pyruvic acid-to-lactic acid pathway appeared to be the key process, with increased conversion of pyruvate to lactate. The researchers suggested that lactate could be a marker of arthritic inflammation.

When the metabolites from the four types of arthritis were compared with each other, there were clear differences which allowed them to be differentiated. For instance, a panel of 13 metabolites differentiated the female RA patients from those with OA. Their combination gave a sensitivity of 85% and a specificity of 81% between the two phenotypes. Similarly, AS was distinguished from gout in male patients by a collection of 16 metabolites, again with good sensitivity and specificity.

If tests could be devised specifically for these sets of metabolites, they could be used to give an early indication of the onset of arthritis and point to the specific type and appropriate treatments. However, the research team were at pains to point out that the number of patients tested should be increased far beyond the current size to confirm the conclusions. Once that takes place, they will have a better idea whether this could lead to a new approach to personalised medicine for arthritic patients.