eBook: Multiplex Technology Guide - 14

plasma samples measured using both platforms
were compared to evaluate capture within each
assay's dynamic range. Bland-Altman plotting elucidated agreement between LMX and MSD for cytokine measurement.21 Heat maps colorimetrically
presented respective protein expression levels of
shared analyte concentrations in plasma samples
when measured by LMX vs. MSD.

Results
Standard Curve Reproducibility: Both assays provided good standard curves for all analytes, as anticipated from the performance data in each user manual.14,15 Inter-assay coefficient of variation (CV) was
generally ≥15% only at the very lowest standard
concentrations. With LMX, the CV of the lowest standard point (above zero) was >25% in 8/16 (50%) of
common analytes; with MSD, the CV of the lowest
standard was >25% in 4/16 (25%) of common analytes (not shown). The range of average CVs was
2.4-13.9% with MSD QCs (high, medium, and lowend concentrations). With LMX, (high and low-end
concentration QCs only), the average CV range was
1.9-18.2%, with 3/16 (19%) analytes registering a
CV >15% at the low-end concentration for two cytokines (IL-7 and MIP-1α) and at both QC concentrations for one analyte (IL-10).

Time and Motion Study: We compared total time
and labor time (including operator hands-on and/or
vigilant observation time) to evaluate the 16 overlapping analytes between both assays. Time studies
were performed over four days by the same technologist, with at least two independent experiments
evaluated for each assay platform. Experimental
steps were timed by an independent impartial observer from a third-party healthcare consulting firm
(Nexus, Plano, TX).

Analytical Limits of Detection and Quantification,
and Dynamic Range: The LoD and LLoQ were determined for all 16 shared cytokines using both LMX
and MSD instruments and each kit's unique recombinant standard proteins (Table 1). The dynamic
ranges, defined as the spread between the LLoQ and
highest standard curve point (ULoQ), are tabulated
in Table 2 and illustrated in Figure 1 for both systems.
The LMX assay had higher ULoQs for 13 of 16 analytes, a slightly lower ULoQ for IL-17, and essentially
identical ULoQs for IL-6 and MIP-1α compared to the
MSD assay. The MSD platform had lower analytical
LLoQs than LMX for 14 of 16 shared cytokines. For
IL-17, the LLoQ was lower in the LMX assay than with
MSD, and the LLoQs for IL-6 were similar between
platforms. In two instances (IL-4 and IL-10), the high
and low quantification limits differed by more than
an order of magnitude between both systems. This
suggested that the MSD assay would demonstrate
better performance than the LMX assay for detecting

Statistical Analysis: All LMX data (median fluorescence intensity) was converted to analyte concentrations (pg/mL) by back-calculations against standard
curves using a multiparametric fit, using Bio-Plex
Manager software v6.2 (Bio-Rad, Hercules, CA). MSD
data (electrochemiluminescence values) was converted to tangible analyte concentrations by curve
fitting and back-calculation using MSD Discovery
workbench software v3.0/4.0. Data was analyzed using R statistical software v3.3.1 (r-project.org) and
Excel v16.0 (Microsoft Inc., Redmond, WA).
Table 1. Demographics of plasma donors.
Parameter

Healthy

Diabetic

High PCT

N

20

21

21

Gender (Male/Female)

5/15

11/10

12/9

Age, Mean ± SD (Years)

43.1 ± 9.8

47.0 ± 15.7

66.2 ± 17.6

Age Range (Years)

26-60

20-79

23-87

14
http://www.r-project.org
eBook: Multiplex Technology Guide

Table of Contents for the Digital Edition of eBook: Multiplex Technology Guide
