100 R. Murray et al.: J Extra Corpor Technol 2025, 57, 96 – 104
Table
2. Pre and post MCS cannulation biomarker kinetics.
Biomarker |
Pre-cannulation, N = 14 1 |
Post-cannulation, N = 14 1 |
P-value 2 |
Trem 1( ng / L) |
446( 122, 245, 549, 1354) [ 488, 339 ] |
371( 183, 264, 463, 898) [ 390, 185 ] |
0.6 |
Presepsin( pg / mL) |
39,630( 4477, 25,142, 73,225, 318,592) |
46,498( 6267, 24,368, 142,630, 1,059,987) |
0.5 |
|
[ 73,539, 87,577 ] |
[ 157,420, 285,461 ] |
|
CRP( mg / mL) |
31( 1, 1, 112, 676) [ 111, 196 ] |
61( 19, 44, 187, 632) [ 131, 167 ] |
0.2 |
PCT( ng / mL) |
0( 0, 0, 1, 94) [ 7, 25 ] |
1( 0, 0, 2, 57) [ 5, 15 ] |
0.4 |
Legend: 1 Median( Minimum, Interquartile Range, Maximum) [ Mean, Standard Deviation ], 2 Wilcoxon rank sum exact test; Wilcoxon rank sum test. Soluble Triggering Receptor expressed on Myeloid cells( sTREM-1), C-Reactive Protein( CRP). N is 14 as 4 patients either developed an infection in the first 48 h of mechanical circulatory support or did not have a pre and post cannulation residual sample available for analysis. Due to the use of residual samples, there were some instances of patients missing samples.
Figure 1. Kinetics of biomarkers while on mechanical circulatory support. Legend: Kinetics of each biomarker during uninfected time periods for the first five days of mechanical circulatory support( Lines indicate median values; boxes indicate 25 – 75th percentiles; and whiskers indicate the range). Soluble Triggering Receptor expressed on Myeloid cells( sTREM-1), C-Reactive Protein( CRP), Procalcitonin( PCT).
are presented in Table 4. Using linear mixed effects modeling we compared the average values for each of the four biomarkers during their post cannulation uninfected periods to the average values during infected periods, ie. the 48 h before infection. On average during periods of infection, Presepsin was 41 % lower( 51,462 – 30,188 pg / mL)( P = 0.001) and procalcitonin was 51 % lower( 0.77 – 0.38 ng / mL)( P < 0.001) while there was no difference in sTrem-1 or CRP compared to non-infected periods. To further examine the response of each biomarker to the development of new infection we compared the average biomarker values during the uninfected period(> 48 h in advance of the infection) to the biomarker values on the day the positive culture was obtained. In this analysis, only Procalcitonin reached significance and was 37 % lower
( 0.98 – 0.62 ng / mL)( P = 0.04). There was no difference found in this analysis for sTREM-1, CRP or Presepsin.
In an exploratory fashion we utilized ROC analysis to determine the optimal cut-point for each of the four biomarkers. The optimal cut-point for sTREM-1 was determined to be 431 ng / L, with an area under the receiver operating characteristic curve( AUC) of 0.56, with a sensitivity of 0.6 and a specificity of 0.54. The optimal cut-point for Presepsin was 64,195 pg / mL with an AUC of 0.553, with a sensitivity of 0.8 and a specificity of 0.36. The optimal cut-point for CRP was 220 mg / ml with an AUC of 0.522, with a sensitivity of 0.39 and a specificity of 0.78. The optimal cut-point for procalcitonin was 0.478 ng / mL with an AUC of 0.618, with a sensitivity of 0.67 and a specificity of 0.59.