Percutaneous vertebroplasty is an efficient procedure to treat pain due to osteoporotic vertebral compression fractures. However, refracture of cemented vertebrae occurs occasionally after vertebroplasty. It is unclear whether such fractures are procedure-related or part of the natural course of osteoporosis. The effect of potentially important covariates on refracture risk in cemented vertebrae has not been evaluated previously. We retrospectively analyzed the incidence and possible causative mechanism of refracture in patients who had received only one vertebroplasty for a single level of vertebral compression fracture. We assessed the following covariates: age, sex, body weight, height, lumbar spine bone mineral density, treated vertebral level, pre-existing untreated vertebral compression fracture, and gas-containing vertebrae before treatment. Surgical variables, including surgical approach, cement injected, and anterior vertebral height restoration, were also analyzed. Antiosteoporotic treatment after surgery was recorded. Multiple logistic regression analysis was used to determine the relative risk of refractures of cemented vertebrae. Over all, 98 patients were evaluated with a mean follow-up of 26.9 ± 12.4 months (range, 7-55 months). We identified 62 refractures and the mean loss of anterior vertebral height was 13.3% (range 3.2-40.3%). The greater the anterior vertebral height obtained from vertebroplasty, the greater the risk of refracture occurring (P \ 0.01). Gas-containing vertebrae were also prone to refracture after the procedure (P = 0.01). Anti-osteoporotic treatment was of borderline significance between refractured and non-refractured vertebrae (P = 0.07). Only restoration of anterior vertebral height was positively associated with refracture during the follow-ups (P \ 0.01). In conclusion, refractures of cemented vertebrae after vertebroplasty occurred in 63% of osteoporotic patients. Significant anterior vertebral height restoration increases the risk of subsequent fracture in cemented vertebrae.

The binding of autoantibodies (autoAbs) to interferon (IFN)-γ in people with mycobacterial diseases has become an emerging medical concern. Many patients display specific human leukocyte antigen (HLA) class II haplotypes, which suggests that a common T cell-dependent and B cell-dependent mechanism might underlie the production of specific anti-IFN-γ autoAbs. We show here that these autoAbs target a major epitope (amino acids 121-131, designated position (P)121-131) in a region crucial for IFN-γ receptor (IFN-γR) activation to impair IFN-γ-mediated activities. The amino acid sequence of this epitope is highly homologous to a stretch in the Noc2 protein of Aspergillus spp., which was cross-reactive with autoAbs from patients. Rats immunized with Aspergillus Noc2 developed antibodies that reacted with human IFN-γ. We generated an epitope-erased variant of IFN-γ (EE-IFN-γ), in which the major neutralizing epitope region was altered. The binding affinity of anti-IFN-γ autoAbs for EE-IFN-γ was reduced by about 40%, as compared to that for IFN-γ1-131. Moreover, EE-IFN-γ activated the IFN-γR downstream signaling pathway ex vivo, irrespectively of anti-IFN-γ autoAbs. In conclusion, we identified a common, crucial B cell epitope that bound to anti-IFN-γ autoAbs in patients, and we propose a molecular-mimicry model for autoAb development. In addition, treatment with EE-IFN-γ might be worth investigating in patients producing anti-IFN-γ autoAbs.

No abstract