In the absence of mutant genes encoding components of the podocyte slit diaphragm, about 30–50 % of children with primary glucocorticoid-resistant focal segmental glomerulosclerosis (FSGS) develop recurrent proteinuria and slowly progressive FSGS lesions following renal transplantation. Recurrence of FSGS in the allograft strongly suggests a circulating factor that disturbs normal podocyte biology. To date, the nature of the circulating factor is unclear, and there is no cure for the recurrent form of FSGS (R-FSGS).

Cultured differentiated human podocytes were exposed to the plasmapheresis effluent or blood plasma samples from pediatric patients with recurrent or primary FSGS; in some cases, podocytes were pre-incubated with specific antibodies to block the tumor necrosis factor-alpha (TNFα) signaling pathway. Integrity of focal adhesion complexes and actin cytoskeleton were investigated by immunofluorescent microscopy.

Plasmapheresis effluent from an R-FSGS child or fresh plasma from two children with primary FSGS rapidly disturbed the cytoskeleton of normal human podocytes in vitro. Plasma from a child with R-FSGS also activated β3 integrin and dispersed focal adhesion complexes. The effects were reversed by pre-incubation with antibodies against TNFα or either of the two TNFα receptors. When our patient with R-FSGS became resistant to plasmapheresis, we initiated treatment with twice weekly etanercept injections and then infliximab. Within 3 weeks of regular anti-TNFα therapy, the patient achieved sustained partial remission of proteinuria, allowing us to wean her off plasmapheresis completely.

We suggest that in some FSGS patients, disruption of the podocyte cytoskeleton and β3 integrin-mediated podocyte attachment are driven by the TNFα pathway.