The detection rates of serum KD-specific MAMPs at each study.

At the 2nd and 3rd studies, both in vivo biofilms and serum samples were simultaneously collected. In vivo biofilms samples were searched for MAMPs common to those in serum samples by LC-MS and MS/MS analyses. SBA: simultaneous biofilm analysis, *: overall % positive (overall positive numbers). DC samples at the 1st (n = 41), 2nd (n = 30) and 3rd (n = 30) study periods were all negative for all KD MAMPs. The detection rates of KD-specific serum MAMPs between KD samples (n = 43) and DC samples (n = 41) at the 1st study showed statistically significant differences at m/z 1531.8, m/z 1414.3, m/z 790.9, m/z 695.0, and overall (P<0.0001), but not at m/z 779.8 (P = 0.1164) by Fisher's exact test. The detection rates between SBA period KD samples (n = 12) and DC samples (n = 30) at the 2nd study showed statistically significant differences at m/z 906.8 (P = 0.0044), m/z 695.0 (P = 0.0002) and overall (P<0.0001), but not at m/z 1171.4 (P = 0.0767) and m/z 1169.4 (P = 0.0767). The detection rates between SBA period KD samples (n = 11) and DC samples (n = 30) at the 3rd study showed statistically significant differences at all 3 molecules and overall (P<0.0001). The detection rates of serum KD-specific MAMPs at each study.