The paper is not available open access, which is a real pity because one has to read it carefully to understand it. I am quite frankly surprised that it's not available online since the work is supported by several NIH grants. Anyway, while I accessed the paper using my online resources, I can only provide you with the pubmed abstract.
The gist of the paper is as follows:Berthoux et al from France selected 97 patients from a larger cohort. They classified the selected patients according to their absolute renal risk (ARR) for progression to dialysis or death (0, very low; 1, low; 2, high; 3, very high). They then analyzed serum samples obtained at diagnosis in the 97 patients for autoantigen and autoantibodies from 97 patients with IgAN. They also analyzed samples from controls comprising 30 healthy volunteers and 30 patients with non-IgAN disease. The mean follow-up was 13.8 years. They report that mean serum levels of total autoantigen, normalized IgG autoantibody, and total IgA autoantibody were significantly higher in patients than in the combined controls (all P less than 0.01). They also reported that the levels of IgG but not IgA correlated with worse clinical outcomes - in Cox regression and Kaplan–Meier analyses, IgG autoantibody levels greater than 1.33 predicted dialysis or death (both P less than 0.01).
The authors discussed the limitations of their study, including that they could not infer a direct causal relationship between the serum levels of either autoantigen or autoantibodies and disease progression, and that a longitudinal study would be required. They also cite a methodlogic flaw in that there was a variation in the length of the intervals between clinical onset and diagnosis by biopsy - "with a greater interval for IgAN subgroups with an ARR of 2 or 3, indicating that the diagnosis was established by biopsy likely later in the clinical course."
I can add several other important limitations - for example:
The sample size was rather small. We are also not told how the 97 subjects were selected from the larger cohort. The magnitude of the difference in the mean levels for IgA or IgG levels between patients with the different degrees of severity was not very great. Further, the authors take a rather unusual approach toward presenting their data. In Table 1, it would have been useful to have provided the clinical data on patients with the different levels of auto antigen, IgG and IgA (expressed perhaps categorically) rather than show us differences according to disease severity. Of course, we know that patients with greater disease severity are likely to have worse hypertension and kidney function and more proteinuria, the question is whether the patients with higher IgA or IgG levels had worse hypertension, kidney function and proteinuria - could these factors provided a better explanation for poorer outcome. Besides this, it would have been important for the authors to have provided information on how they built their regression models and what factors they adjusted for.
Still, taking all of the limitations into account, it was interesting that the patients with more severe IgAN (ie those who progressed) had higher IgG and IgA levels. What I didn't like, frankly, was the lyrical leap from a well written discussion that did not claim that these parameters are able to predict severity to a Kidney News article that these tests could provide a "warning sign" for patients. Hardly something we should be saying until we know with some degree of confidence that these tests are truly predictive (and in fact have some sense (sensitivity, specificity, positive and negative predictive value) of how predictive they are.
