Component Description
Monoclonal gammopathy of
undetermined significance (MGUS) is a precursor condition to multiple myeloma
and related plasma cell disorders. Prior research using samples from the
National Health and Nutrition Examination Survey (NHANES) has demonstrated the
prevalence of MGUS in the U.S. population and its variation by age, race, and
sex (Landgren, et. al., 2014; Landgren, et. al., 2017).
Mass spectrometry-based assays
have been proposed and applied as high-sensitivity methods for the detection of
monoclonal gammopathies (El-Khoury, et. al., 2022). Serum samples were analyzed
using a high-resolution mass spectrometry platform capable of detecting
multiple immunoglobulin protein peaks within a single sample.
Eligible Sample
Examined participants aged 40 years and over were eligible.
Description of Laboratory Methodology
M-protein
(monoclonal protein) studies with matrix-assisted laser desorption-ionization
time of flight mass spectrometry (MALDI-TOF MS) were carried out using the
automated EXENT system (The Binding Site Group Ltd. Birmingham, UK) and
immunoglobulin isotype assay. The immunoglobulin assay uses sheep polyclonal
antibodies (specific for anti-human IgG, IgA, IgM, total κ, and total λ light
chains) separately immobilized onto paramagnetic microparticles. Immunoglobulin
purification was performed on the system’s automated liquid handler. Briefly,
participant samples were equally divided into 5 aliquots, each individually
incubated with isotype-specific beads. The beads were then washed using
magnetic precipitation and buffer exchanges and treated with 20mM Tris [2-carboxyethyl]
phosphine (TCEP) in 5% (v/v) ac/etic acid to reduce immunoglobulin heavy and
light chain disulfide bonds. Eluates were subsequently mixed with
α-cyano-4-hydroxycinnamic acid (HCCA) matrix, spotted onto MALDI plates, and
allowed to co-crystallize on the plate surface. Sample spots were analyzed on a
MALDI-TOF SirusⓇ (Bruker, CA, U.S.). LC mass spectra were
acquired in positive ion mode covering the m/z range of 5,000 to 32,000, which
includes the singly charged (+1, m/z 23,330 to 24,650) and doubly charged (+2,
m/z 11,168 to 12,401) ions. LC m/z distributions from each reaction (IgG, IgA,
IgM, total κ, and total λ) were simultaneously interrogated using the EXENT
software. Monoclonal peak picking was achieved following a proprietary
algorithm trained against clinical data developed by Indigo Bioautomation.
Briefly, after finding the local y-maxima a spline model was used to assess the
goodness of fit. A spline model is necessary for MALDI-TOF data obtained using
HCCA matrix to accommodate the matrix adduct peak present at a regularly defined
distance from the monoclonal peak. Polyclonal peak modeling was achieved using
a Gaussian fit. The area under the curve for the +2 charge state was calculated
using the trapezoidal rule to account for the area above the baseline.
Quantification of total immunoglobulin levels was determined by turbidimetry on
an Optilite® instrument (The Binding Site Group Ltd.).
Laboratory Quality Assurance and Monitoring
The
analytical performance of the EXENT system has previously been verified for
linearity, precision, reference interval, interference, and stability. EXENT’s
limit of quantification (LoQ) is 15 mg/L for all specificities, with a limit of
detection (LoD) of <5 mg/L dependent on the polyclonal background. Published
Hevylite® (The Binding Site Group Ltd.) reference intervals have been verified
for each specificity. The assay demonstrates acceptable 20-day total precision
values for polyclonal samples (<7%) and for two monoclonal samples (~1 g/L
and 10 g/L; <15% and <10%, respectively), with between laboratory
precision <8% for all samples. The assay is linear for all three major
isotypes (IgG, IgA, or IgM) across a 0.015 to 100 g/L dynamic range. The Thermo
Fisher Scientific’s EXENT System received clearance from the U.S. Food and Drug
Administration (FDA) to aid in the diagnosis of multiple myeloma and other
conditions.
Data Processing and Editing
Data were received after all
analyses were complete. The data were reviewed. Incomplete data or improbable
values were sent to the performing laboratory for confirmation.
Specific variables
and edits:
SSMSDIAG
Monoclonal
gammopathy classification (SSMSDIAG) was coded as 1 (MGUS) when max
concentration of any peak (SSMSMAX) ≥0.2 g/L, coded as 2 (MGIP; monoclonal
gammopathy of indeterminate potential) when SSMSMAX ≥0.015 g/L and SSMSMAX <
0.2 g/L, coded as 3 (negative) when number of detected peaks (SSMSPN) = 0 or when
all detected peaks were below the lower limit of quantification (<0.015 g/L)
or appeared only as light chains (i.e. kappa or lambda) with no matched heavy
chain, and coded as Missing when no specimen was available for testing.
SSMSMAX
Max concentration
of any peak (SSMSMAX) is reported as concentration of the peak with the highest
measured concentration. If there were no detected peaks, or if all detected peaks
were below the LLoQ (<0.015 g/L), or if all peaks appeared only as light
chains (i.e. kappa or lambda) with no matched heavy chain, SSMSMAX is coded as
Missing.
SSMSPN
Number of detected
peaks (SSMSPN) is reported as the total number of detected monoclonal protein (M-protein)
peaks. Any detected peak, regardless of concentration, is included in this
count.
SSMSP1T-SSMSPBLC: Information for up to 11 M-protein peaks is included. Each detected peak is characterized by its mass-to-charge ratio (m/z), immunoglobulin isotype, and, when quantifiable, its concentration. Multiple peaks may be observed per participant, reflecting the presence of one or more immunoglobulin clones. Four variables are provided for each M-protein peak, including: peak immunoglobulin isotype (SSMSP_T), mass-to-charge ratio of peak (SSMSP_MZ), concentration of peak in g/L (SSMSP_C), and comment code (SSMSP_LC). For each variable, the peak is denoted by a corresponding letter or number (ex., SSMSP1T = peak isotype for peak #1). Peak numbers are assigned in descending order of concentration, when quantifiable. Peaks are assigned the following letters or numbers: Peak #1 = 1; Peak #2 = 2; Peak #3 = 3; Peak #4 = 4; Peak #5 = 5; Peak #6 = 6; Peak #7 = 7; Peak #8 = 8; Peak #9 = 9; Peak #10 = A; Peak #11 = B. In all cases of a detected peak, a peak immunoglobulin isotype is provided (SSMSP_T). Concentrations (SSMSP_C) are rounded to three significant figures and reported to a maximum of 3 decimal places. The assay allows detection of M-protein peaks below the LLoQ. In such cases, the presence of a peak is recorded, but concentration values are listed as Missing rather than imputed or set to a fixed value; instead, a comment code is provided indicating the concentration is below the LLoQ. For peaks with light chains only (i.e. kappa or lambda) with no matched heavy chain, no concentration is provided. The mass-to-charge ratio for each peak is listed (SSMSP_MZ). In cases where the mass-to-charge ratio has a suspected shift due to post-translational glycosylation, the mass-to-charge ratio may be missing if it is unable to be estimated. A comment code indicates cases of suspected shift in mass-to-charge ratio due to post-translational glycosylation. The comment code (SSMSP_LC) for each peak is assigned as “1” (Below LLoQ) when the concentration of the peak is <0.015 g/L, assigned as “2” (MSL) when post-translation glycosylation results in an altered mass-to-charge ratio, assigned as “3” (Below LLoQ + MSL) when both conditions exist, or not assigned any comment code when neither condition exists.
SSMSIGA, SSMSIGG,
SSMSIGM, SSMSK, SSMSL
These variables
were rounded to two decimal places.
SSMSFLC
Free light chain
ratio (SSMSFLC) was calculated as SSMSK (prior to rounding) divided by SSMSL
(prior to rounding) and rounded to two decimal places.
Analytic Notes
Refer to the 2001-2002 Laboratory Data Overview for
general information on NHANES laboratory data.
There are over 800 laboratory
tests performed on NHANES participants. However, not all participants provided
biospecimens or enough volume for all the tests to be performed. Additionally,
availability of specimens for surplus projects is lower than for other
laboratory tests performed on NHANES participants. The specimen availability
can also vary by age or other population characteristics. Analysts should
evaluate the extent of missing data in the dataset related to the outcome of
interest as well as any predictor variables used in the analyses to determine
whether additional re-weighting for item non-response is necessary.
Please refer to the NHANES Analytic Guidelines and
the on-line NHANES Tutorial for details on the use of sample weights and analytic issues.
Demographic,
Sample Weights, and Other Related Variables
The analysis of NHANES laboratory
data must be conducted using the appropriate survey design and demographic
variables. The NHANES 2001-2002 Demographics File contains
demographic data, health indicators, and other related information collected
during household interviews as well as the sample design variables and
examination weights (WTMEC2YR and WTMEC4YR). The recommended procedure for
variance estimation requires use of stratum and PSU variables (SDMVSTRA and
SDMVPSU, respectively) in the demographic data file.
This laboratory data file can be
linked to the other NHANES data files using the unique survey participant
identifier (i.e., SEQN).
Quantification Limits and
Comment Codes
Please refer to the above Data
Processing and Editing section for more information on variable definitions and
coding relevant to analyses.