Evaluation of potential interference of autoantibodies with measurements of survivin by ELISA. (A) Twenty randomly chosen serum samples with known levels of survivin (Surv), rheumatoid factor (RF) and autoantibodies to cyclic citrullinated peptides (aCCPs) were used to prepare a survivin-positive pool (pool A), a RF-positive pool (pool B), a RF- and aCCP-positive pool (pool C) and a RF- and aCCP-negative pool (pool D). Serial dilution of pool A showed 98% recovery of survivin levels with sample dilution. (B) Pool A (survivin-positive) was diluted with pool B (RF-positive). The measurements of survivin were not affected by these increasing concentrations of RF. (C) A spike recovery procedure whereby pool D (RF- and aCCP-negative), pool B (RF–positive and aCCP-negative) and pool C (RF- and aCCP-positive) were spiked with recombinant survivin. The measurements of survivin showed 85% to 100% recovery, which was similar for all tested pools.