Statistics from Altmetric.com
We were interested in the study by Berthelot and colleagues, in which they studied the predictive value of the antiperinuclear factor (APF) for rheumatoid arthritis.1
The authors concluded that APFs are useful in the diagnosis of early rheumatoid arthritis. They mentioned that recognition of the true value of APF has long been hindered by methodical errors. Serum samples were only diluted 1:5 or 1:10, even though the APF titre can reach 1:20 000 in rheumatoid artritis serum samples and is usually above 1:200.
We would like to make some comments especially concerning the method they used. The method used differs not only in the dilution of the serum samples, for screening 1:100 and if positive they were further diluted 1:200, 1:500, 1:1000 to determine the end point titre, but it differs also in the way of regarding the positivity of the tests. They refer to their study published in 1990, in which a dilution of 1:80 is used. In 1990 Westgeest et al described for the first time the influence of serum dilution on findings of the APF prevalence in rheumatoid arthritis (RA).2 So authors of the study on an interlaboratory variability test for RA (1993) from the representing laboratories in Amsterdam, Brest, Gent, and Nijmegen came together to discuss their different methods for the APF determination and to reach a standard protocol. This protocol was used for a quantitative assessment of the method, using the WHO standard rheumatoid artritis serum as a reference serum. The results of coded serum samples from five separate patients were blindly tested in the five laboratories. The methods used were not harmonised. Four laboratories defined the titre of a serum by the highest dilution, which gave a clearly positive result, while the laboratory in Brest, to which Berthelot and colleagues also refer, also considered the percentage of cells that were stained.3 The results of the five laboratories showed only small interlaboratory variations. The interlaboratory variations were again reduced by comparison with the results with the WHO standard rheumatoid arthritis serum and thereby expressing the results in IU instead of litres, this study is not mentioned in Dr Berthelot’s study.1 So linked to the uniformity of the results, we think that the value of the APF is not hindered by the method.
In our longitudinal evaluation during methotrexate and azathioprine treatment for RA, the highest APF titre was 1:640.4 In the study by Berthelot et al the APF titre can reach 1:20 000 in rheumatoid arthritis serum samples.
Did Berthelot and colleagues use another method than that they used in the intervariability study?
We gratefully acknowledge Dr Boerbooms for her valuable comments on our antiperinuclear factor (APF) paper.1-1 Her statement is correct. Five European groups set up a consensus study on the interlaboratory variability of the APF test in 1993.1-2Despite the use of different cells, conjugates, and criteria for positivity, their results were comparable for the five serum samples tested. Given that there is a risk that any artefact could be mistaken for a stained granule, we feel, however, that it is useful to take into account the proportion of stained cells,1-3 as suggested by others.1-4 Similarly Dr Westgeest and Boerbooms underlined themselves the importance of serum dilution,1-5 the same year when Youinou et al proposed to increase it up to 1:80 instead of 1:5.1-4 We recently identified another source of poor results (that is, to delay the test several days after fixation of the slides).1-6 Regarding our much better sensitivity, Dr Boerbooms is also aware that the suitability of particular buccal mucosa cells is unpredictable. The possibility therefore exists that donors vary from one country or laboratory to another, and actually more than 70% of our donor’s cells exhibit large granules. We fully agree with Dr Boerbooms that the APF test warrants being used on a regular basis. We recently came across the literature and believe that the APF test has been underevaluated. To conclude, we thank Dr Boerbooms for the opportunity to debate these technical details, although we quite respect her choice of an older but still valuable method: for instance Westgeest and Boerbooms could confirm the findings of Katahaa1-7 that APF can also be found in some infectious mononucleosis serum samples, and suggested that Epstein-Barr virus might be a co-factor in the breakdown of tolerance towards profilaggrin.1-8 Although direct evidence for this assumption is still lacking, our recent findings of a strong T cell response towards Epstein-Barr virus lytic cycle antigens in RA joints might indicate that their hypothesis was correct.1-9
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.