Clin Chem Lab Med 2016; aop
Maria Alice V. Willrich*, Paula M. Ladwig, Bruna D. Andreguetto, David R. Barnidge, David L. Murray, Jerry A. Katzmann and Melissa R. Snyder
Monoclonal antibody therapeutics as potential interferences on protein electrophoresis and immunofixation DOI 10.1515/cclm-2015-1023 Received October 19, 2015; accepted December 13, 2015
Abstract Background: The use of therapeutic recombinant monoclonal antibodies (mAbs) has triggered concerns of misdiagnosis of a plasma cell dyscrasia in treated patients. The purpose of this study is to determine if infliximab (INF), adalimumab (ADA), eculizumab (ECU), vedolizumab (VEDO), and rituximab (RITU) are detected as monoclonal proteins by serum protein electrophoresis (SPEP) and immunofixation electrophoresis (IFE). Methods: Pooled normal sera were spiked with various concentrations (ranging from trough to peak) of INF, ADA, ECU, VEDO and RITU. The peak concentration for VEDO and RITU was also added to samples with known monoclonal gammopathies. All samples were analyzed by SPEP (Helena Laboratories) and IFE (Sebia); sera containing peak concentrations of mAbs were reflexed to electrospray-time-of-flight mass spectrometry (AbSciex Triple TOF 5600) for the intact light chain monoclonal immunoglobulin rapid accurate mass measurement (miRAMM). Results: For all mAbs tested, no quantifiable M-spikes were observed by SPEP at any concentration analyzed. Small γ fraction abnormalities were noted on SPEP for VEDO at 300 μg/mL and RITU at 400 μg/mL, with identification of small IgG κ proteins on IFE. Using miRAMM for peak samples, therapeutic mAbs light chain accurate masses
*Corresponding author: Maria Alice V. Willrich, PhD, Division of Clinical Biochemistry and Immunology, Department of Laboratory Medicine and Pathology, Mayo Clinic, 200 1st St SW, Rochester, MN 55905, USA, Phone: +507-266-4909, Fax: +507-266-4088, E-mail: [email protected] Paula M. Ladwig, David R. Barnidge, David L. Murray, Jerry A. Katzmann and Melissa R. Snyder: Division of Clinical Biochemistry and Immunology, Department of Laboratory Medicine and Pathology, Mayo Clinic, Rochester, MN, USA Bruna D. Andreguetto: From the Department of Clinical Pathology at Universidade de Campinas, Campinas, SP, Brazil
were identified above the polyclonal background and were distinct from endogenous monoclonal gammopathies. Conclusions: MAbs should not be easily confounded with plasma cell dyscrasias in patients undergoing therapy except when a SPEP and IFE are performed within a couple of days from infusion (peak). In ambiguous cases the use of the miRAMM technology could precisely identify the therapeutic mAb distinct from any endogenous monoclonal protein. Keywords: immunofixation; monoclonal antibody therapies; monoclonal immunoglobulin Rapid Accurate Mass Measurement (miRAMM); protein electrophoresis.
Introduction Therapeutic monoclonal antibodies (mAbs) have rapidly become a clinically important drug class, as the efficacy of their use for treatment of a wide variety of clinical indications becomes proven. There are applications that include autoimmunity and inflammation, cancer, organ transplantation, cardiovascular disease, infectious diseases and ophthalmological diseases [1]. Human immunoglobulins (Ig) are composed of two identical heavy chains and two identical light chains, held together by disulfide bonds. Although there are five types of immunoglobulins defined by their heavy chains (IgG, IgA, IgM, IgD, and IgE) and two types of light chains (κ and λ), the vast majority of the therapeutic mAbs currently on the market are of the IgG κ isotype. The IgG family of antibodies may be further divided, based on the structure of their heavy chains, into four subclasses: IgG1, IgG2, IgG3, and IgG4. Structural differences among IgG heavy chains lead to differences in binding of the subclasses to Fc receptors and, consequently, to subclass- specific differences in processes mediated by Fc receptors [2]. Although most therapeutic mAbs are IgG1, there are many IgG2 and IgG4 currently approved as well. The initial therapeutic mAbs were generated from mouse and rat hybridomas. These first-generation
Brought to you by | Washington University in St. Louis Authenticated Download Date | 2/24/16 12:49 PM
TNF-α, Tumor necrosis factor-α; I.V., intravenous; S.C., subcutaneous; IBD, inflammatory bowel disease; RA, rheumatoid arthritis; aHUS, atypical hemolytic uremic syndrome; PNH, p aroxysmal nocturnal hemoglobinuria; NHL, non-Hodgkin’s lymphoma; CLL, chronic lymphocytic leukemia; GPA, granulomatosis with polyangiitis; MPA; microscopic polyangiitis. aHalf-life obtained from drug package insert or prescribing information.
IBD, RA IBD, RA aHUS, PNH IBD NHL, CLL, GPA, MPA 5–10 mg/kg every 8 weeks 40–80 mg every other week 300–1200 mg every other week 300 mg every 8 weeks 500 mg per adult I.V. S.C. I.V. I.V. I.V. 8–10 days 14 days 11.3 days 25 days 22 days Chimeric IgG1 κ Human IgG1 κ Humanized IgG2/IgG4 κ Humanized IgG1 κ Chimeric IgG1 κ TNF-α TNF-α Complement factor C5 α4β7 integrin CD 20
Administration dose Administration route Average Half-lifea Structure Target molecule
Brought to you by | Washington University in St. Louis Authenticated Download Date | 2/24/16 12:49 PM
Infliximab Adalimumab Eculizumab Vedolizumab Rituximab
INF, ADA, VEDO and RITU were purchased from the Mayo Clinic pharmacy and reconstituted per manufacturer instructions. ECU was supplied in a ready for use formulation. Reagents for SPEP were purchased from Helena Laboratories (Beaumont, TX, USA) and for IFE from Sebia
Materials and methods
Table 1: Characteristics of the studied monoclonal antibodies.
antibodies, named -omab, found only limited success in the clinic because of their short half-lives and high immunogenicity. A number of approaches have been developed to humanize rodent antibodies, from development of chimeric antibodies (–ximab), where constant and variable regions are derived from human and rodent IgG, respectively, to humanized antibodies, where 90%–95% of the antibody is composed of sequence derived from human IgG with only the complementarity determining regions (CDR) being of murine source (–zumab), to fully human antibodies (–umab). Serum protein electrophoresis (SPEP) and immunofixation (IFE) are an important part of the diagnostic work-up for patients with suspected plasma cell dyscrasias. However, SPEP/IFE may also be ordered for a number of other indications, including autoimmune/infectious diseases (polyclonal hypergammaglobulinemia), renal dysfunction (nephrotic proteinuria), and primary immunodeficiency (hypogammaglobulinemia). For many of these diseases, treatment with a therapeutic mAb is becoming common. This raises the question of whether a therapeutic mAb could be mis-interpreted as an endogenous M-protein on SPEP or IFE, potentially triggering unnecessary additional testing and clinical evaluation. A few reports have described the detection of mAb therapeutics on SPEP or IFE in patients under treatment, such as the detection of siltuximab, against interleukin-6 [3] and ofatumumab, an anti-CD20 antibody [4]. Siltuximab is a chimeric (30% murine/70% human) IgG κ mAb. At the time of the reported interference on SPEP, siltuximab was being used in clinical trials for the treatment of refractory multiple myeloma in combination with dexamethasone and was identified in a patient with history of IgD κ myeloma. The purpose of this study was to determine if mAbs used for chronic inflammatory conditions [INF, adalimumab (ADA), eculizumab (ECU), vedolizumab (VEDO), and rituximab (RITU)] (Table 1) could be detected on SPEP or IFE at trough, mid-dose and peak therapeutic concentrations. Secondly the use of time-of-flight (TOF) mass spectrometry was investigated as a way to accurately distinguish between therapeutic and endogenous monoclonal proteins.
Conditions where it is commonly prescribed
2 Willrich et al.: Monoclonal antibody therapeutics as interferences on SPEP and IFE
Willrich et al.: Monoclonal antibody therapeutics as interferences on SPEP and IFE 3 (Norcross, GA, USA). Reagents for mass spectrometry included ammonium bicarbonate, dithiothreitol and formic acid (Sigma-Aldrich Inc., Saint Louis, MO, USA). Melon Gel was purchased from Thermo Fisher (Waltham, MA, USA) and water, 2-propanol and acetonitrile were from Honeywell Burdick and Jackson (Muskegon, MI, USA). Residual normal sera with no abnormalities on SPEP and IFE were pooled and spiked with three different concentrations for each mAb, ranging from trough to peak, as described in Table 2. Concentrations were defined based on studies with in-house serial sample collections from patients undergoing therapy with these mAbs. In order to verify the experiments with normal serum spiked with mAbs, residual serum samples from patients undergoing therapy with INF, ADA, ECU and VEDO were used (n = 4). Concentrations of the mAbs were measured using tryptic peptide mass spectrometry methods (INF and VEDO) [5, 6], a commercial immunoassay for ADA (Esoterix Laboratories, Calabasas, CA, USA) and ELISA for ECU [7]. All samples were analyzed using standard operating procedures from the clinical laboratory for SPEP (Helena Laboratories) and IFE (Sebia) [8]. A non-quantifiable SPEP abnormality, denoted by term “small”, is defined by the laboratory as an observable electrophoretic restriction or asymmetry that is too small to fractionate as an M-spike. A third group of de-identified residual waste serum samples with small γ migrating abnormalities and subsequent monoclonal IgG κ’s identified on IFE (n = 10) was collected and used for comparison to a fourth group of residual samples without SPEP abnormalities (n = 10) spiked with 500 μg/mL of INF. Finally, two samples with known IgG κ monoclonal gammopathies and quantifiable M-spikes were spiked with peak concentrations of VEDO (300 μg/mL) or RITU (400 μg/mL). Serum samples were analyzed using electrospray-time-of-flight mass spectrometry (AbSciex Triple TOF 5600) for the intact light chain using the monoclonal immunoglobulin Rapid Accurate Mass Measurement (miRAMM) method [9]. Briefly, this method utilizes microflow liquid chromatography electrospray-time-of-flight mass spectrometry to identify the accurate molecular mass of the light chain portion of a monoclonal immunoglobulin. Sample preparation included enrichment of IgG from serum samples using Melon Gel resin (Thermo Fisher). 20 μL of serum was added to 200 μL of resin and mixed for 5 min on an orbital shaker. 20 μL of the supernatant was added to 20 μL of 50 mM ammonium bicarbonate and 10 μL of 200 mM dithiothreitol (DTT) to release light chains from the heavy chains. The mixture was incubated for 15 min at 55 °C before injection onto an Eksigent Ekspert 200 microLC system (Redwood City, CA, USA) in a Poroshell 300SB-C3 analytical column (Agilent) kept in a column heater at 60 °C, using a flow rate of 25 μL/min in a total chromatography run time of 24 min. Mobile phases A (aqueous) and B (90% acetonitrile/10% 2-propanol/0.1% formic acid) were used in a gradient for light chain elution as previously described [10].
The ABSciex tripleTOF 5600 (ABSciex, Framingham, MA, USA) was operated in ESI positive mode with a Turbo V dual-ion source. Source conditions were ISVF 5500/temperature 50 °C/CUR 45/GS1 35/GS2 30. TOF MS scans ranged from m/z of 600–2500 with 200 ms of acquisition time. Analyst TF v1.6 software was used for instrument control and data viewing. The +11 charge state was used for identification of the mAbs in the total ion chromatograms, with the spectra being shown as mass-to-charge (m/z) ratios. The +11 charge state was chosen as it consistently showed a total κ/λ ratio in close agreement to those found by immunonephelometry [11]. The mass spectra of the multiply charged ions were deconvoluted using the Analyst software to a molecular mass characteristic of each light chain in the individual’s immunoglobulin repertoire or the therapeutic mAb. The mass of the intact light chains for each mAbs used as therapeutics were characterized using the pure pharmaceutical preparations (Table 2) diluted in 50 mM ammonium bicarbonate.
Results Pooled normal sera spiked with peak, mid-dose, and trough concentrations for INF, ADA, ECU, VEDO, or RITU were analyzed by routine SPEP and IFE. On SPEP, no quantifiable M-spikes at any concentration of the five therapeutic mAbs were observed. In addition, no small abnormalities were detected on SPEP for any of the “expected” trough or mid-dose concentrations of the therapeutic mAbs. The only small abnormalities noted were for VEDO and RITU at the peak concentrations of 300 μg/mL and 400 μg/mL, respectively (Figure 1). The small band caused by VEDO migrates in the center of the γ fraction and the band from RITU is very cathodal, migrating as a blunt end on the electrophoretic densitogram. On IFE, these small electrophoretic abnormalities seen with VEDO and RITU at the peak concentrations were identified as small IgG κ bands (Figure 2). No abnormalities were identified on IFE for any of the mAbs at trough or mid-dose concentrations (data not shown). A normal serum sample and the samples spiked with peak concentrations of the five therapeutic mAbs were then subjected to miRAMM. The normal serum shows two distinct Gaussian peaks corresponding to the λ and κ polyclonal immunoglobulin repertoires as previously described (Figure 3A) [11]. In the spiked samples, the mAbs used as
Table 2: Typical concentrations observed for monoclonal antibody therapeutics and intact light chain accurate mass measured by miRAMM.
Target therapeutic thresholds at trough
Typical concentrations at mid-term
Typical concentrations at peak
Deconvoluted accurate mass of intact light chains on TOF (+11 charge state m/z)
Infliximab Adalimumab Eculizumab Vedolizumab Rituximab
5 μg/mL 10 μg/mL 50 μg/mL 5 μg/mL 10 μg/mL
30 μg/mL 30 μg/mL 100 μg/mL 30 μg/mL 100 μg/mL
100 μg/mL 100 μg/mL 200 μg/mL 300 μg/mL 400 μg/mL
23,434 Da (2131.4) 23,406 Da (2129.0) 23,130 Da (2103.8) 23,901 Da (2173.9) 23,035 Da (2095.1)
Brought to you by | Washington University in St. Louis Authenticated Download Date | 2/24/16 12:49 PM
4 Willrich et al.: Monoclonal antibody therapeutics as interferences on SPEP and IFE
SPEP 5
30
100
10
inf
30
100
50
100 200
ada
5
ecu
30
300
10
100 400
vedo
ritu
Figure 1: SPEP gels of originally negative SPEP/IFE samples after spiking of mAbs in concentrations expected to be found at trough, middose and peak of therapy regimens. Concentrations shown in μg/mL.
A
B
inf 100
SPEP
G
A
M
ada 100
K
L
SPEP
C
G
A
M
K
L
ecu 200
(µg/mL)
SPEP
D
G
A
M
K
L
E vedo 300
SPEP
G
A
M
ritu 400
K
L
SPEP
G
A
M
K
L
Figure 2: Immunofixation gels. (A) Infliximab at 100 μg/mL was spiked in serum. (B) Adalimumab at 100 μg/mL in serum. (C) Eculizumab at 200 μg/mL. (D) Vedolizumab at 300 μg/mL. (E) Rituximab at 400 μg/mL. Small abnormalities in the γ fraction are noted for vedolizumab and rituximab.
therapies are detectable as a peak above the polyclonal background of endogenous immunoglobulins. Each mAb, with its unique κ light chain, demonstrates a different m/z ratio (Figure 3B–F). The m/z of the +11 charge state as measured by TOF mass spectrometry for each therapeutic mAb in serum was within ±0.5 Da of the mass as determined by measuring the therapeutic mAbs in buffer (Table 2). Samples from patients undergoing therapy with the studied mAbs are shown in Figure 4. SPEP and IFE findings show no abnormalities for INF or ADA. ECU was present in circulation in patient’s serum at 249 μg/mL,
and was additionally observed as a faint small IgG κ on IFE. VEDO was measured at 131 μg/mL, and although not a sharp band on IFE, a dense IgG κ staining is observed. MiRAMM results are presented for the samples and all mAbs accurate mass was identifiable above the polyclonal background, even if not observed on IFE (Figure 4). A sample from a patient undergoing RITU therapy was not available. Next, the ability of miRAMM to distinguish between therapeutic and endogenous mAbs was assessed. Previously characterized serum samples with confirmed small
Brought to you by | Washington University in St. Louis Authenticated Download Date | 2/24/16 12:49 PM
Willrich et al.: Monoclonal antibody therapeutics as interferences on SPEP and IFE 5
A
Normal SPEP/IFE
λ
ada 100 µg/mL 2128.9 Da
vedo 300 µg/mL 2173.9 Da
B
inf 100 µg/mL 2131.4 Da
κ
C
D
ecu 200 µg/mL 2103.7 Da
E
ritu 400 µg/mL 2095.1 Da
F
Figure 3: MiRAMM deconvoluted extracted ion chromatograms. (A) Negative SPEP/IFE serum light chains miRAMM extracted ion chromatogram. The Gaussian distribution of λ and κ light chains is shown. (B) Infliximab at 100 μg/mL in serum shows up as a monoclonal peak above the polyclonal background. (C) Adalimumab at 100 μg/mL. (D) Eculizumab at 200 μg/mL. (E) Vedolizumab at 300 μg/mL. (F) Rituximab at 400 μg/mL. X axis of figures, mass-to-charge ratios (Da). Y axis, intensity (cps).
IgG κ abnormalities migrating in the γ f raction are shown on a representative IFE gel in Figure 5A. Figure 5B, on the other hand, shows IFE results of normal sera spiked with 500 μg/mL of INF. Most of the samples spiked with INF demonstrate a detectable IgG κ monoclonal protein, with the exception of samples #4 and #5, in which an endogenous polyclonal hypergammaglobulinemia (γ fraction of 29 g/L and 15 g/L, respectively) masks the band. The spiked group of samples most likely would have been reported as small IgG κ abnormalities in the γ fraction if these were clinical specimens. The endogenous and spiked samples were then analyzed by miRAMM. The results from the 10 samples with endogenous small IgG κ are shown as overlaid extracted ion chromatograms (+11 charge state). Each sample shows a sharp peak which represents the light chain from the monoclonal IgG κ. However, each light chain has a unique mass, representing diversity across the different monoclonal proteins (Figure 6A). In contrast, for the samples spiked with 500 μg/mL of INF, all of the miRAMM chromatograms show a single peak above the
polyclonal background, corresponding to the INF mass at the +11 charge state. When the spectra are overlaid, all samples spiked with INF show the same mass for the monoclonal κ light chain peak (Figure 6B). The mAb is observed even in the presence of hypergammaglobulinemia, noticeable especially on the red trace chromatogram. Lastly, the ability of miRAMM to identify a therapeutic mAb in the presence of a large endogenous monoclonal protein with a quantifiable M-spike was addressed. M-spikes characterized as IgG κ isotypes were spiked with VEDO at 300 μg/mL or RITU at 400 μg/mL. IFE and miRAMM are shown in Figure 7. The sample spiked with VEDO did not have an additional visible band on IFE. The mAb co-migrates with the M-spike on IFE, however miRAMM was able to separate the endogenous M-spike from the therapeutic mAb based on each clone’s unique mass (Figure 7A and B). The sample spiked with RITU, instead, had an additional band in the γ fraction on IFE. MiRAMM confirmed the additional band was from RITU based on its light chain mass (Figure 7C and D).
Brought to you by | Washington University in St. Louis Authenticated Download Date | 2/24/16 12:49 PM
6 Willrich et al.: Monoclonal antibody therapeutics as interferences on SPEP and IFE
2131.4 Da Patient A Inf 125 µg/mL
2128.8 Da Patient B Ada 36 µg/mL
2103.7 Da Patient C Ecu 249 µg/mL
2173.9 Da Patient D Vedo 131 µg/mL
SPEP G
A
M
K
L
Figure 4: Immunofixation and miRAMM results from residual serum of patients undergoing therapy with different mAbs. Samples were collected at different time-points after therapy, and reflect concentrations usually found at peak (hours-days after dose). Concentrations of the mAbs were measured using tryptic methods and LC-MS/MS (infliximab and vedolizumab), a commercial electrochemiluminescent assay (adalimumab) and ELISA (eculizumab).
A
B
Figure 5: Immunofixation gels. (A) Lanes 1–8, serum samples reported as “small IgG κ in the γ fraction”. Lane 9, IgM κ positive control for immunofixation. (B) Lanes 1–8, serum samples originally negative by immunofixation after spiking of 500 μg/mL of infliximab. On lane 4 it is possible to note hypergammaglobulinemia (original γ fraction of 29 g/L). Lane 9, IgM κ positive control. Note: 10 samples were selected for each group. Only 8 IFE reports shown due to space constrains. Results of omitted samples (9–10) are similar to the reports shown. Brought to you by | Washington University in St. Louis Authenticated Download Date | 2/24/16 12:49 PM
Willrich et al.: Monoclonal antibody therapeutics as interferences on SPEP and IFE 7
A
170
7
160 150 140 130
Intensity, cps
120 110 100
4
3
5
90 80
2
10
70 60 50 40
9a 9b
30 20
6 9c 1 8a
8b
1
2154.906 Da
2
2202.715 Da
3
2174.734 Da
4
2114.526 Da
5
2131.976 Da
6
2150.935 Da
7
2127.539 Da
8
(a) 2128.340 Da (b) 2136.976 Da
9
(a) 2091.937 Da (b) 2097.042 Da (c) 2147.102 Da
10
2130.643 Da
1
2131.411 Da
2
2131.419 Da
3
2131.437 Da
4
2131.451 Da
5
2131.445 Da
6
2131.455 Da
7
2131.451 Da
8
2131.422 Da
9
2131.425 Da
10
2131.428 Da
10 0 2040 2050 2060 2070 2080 2090 2100 2110 2120 2130 2140 2150 2160 2170 2180 2190 2200 2210 2220
Mass/charge, Da
Intensity, cps
B
35 34 33 32 31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
1–10
2020 2030 2040 2050 2060 2070 2080 2090 2100 2110 2120 2130 2140 2150 2160 2170 2180 2190 2200
Mass/charge, Da
Figure 6: MiRAMM results from residual serum samples. (A) miRAMM extracted ion chromatogram showing the +11 charge state zoomed spectra of ten samples reported as small IgG κ in the γ fraction. The table on the right shows the masses of the monoclonal peaks observed above the polyclonal background for each of the samples. Sample 8 had 2 unique clones; sample 9 had 3 clones. (B) miRAMM (+11 charge state) of samples spiked with 500 μg/mL of infliximab. The table on the right shows the measured mass for the monoclonal peaks for all 10 samples.
Discussion Therapeutic mAbs are gaining market share because of proven efficacy in a wide range of clinical disorders. These applications are not only in oncology but also in autoimmune conditions such as rheumatoid arthritis and inflammatory bowel disease. Some of the mAbs are administered at very high doses, especially in the treatment of tumors, usually for a short period of time, as adjuncts to chemo- or radiation therapies. Others, like the ones presented in this study, are used long-term, for treatment of chronic diseases, and are administered at lower doses. In our clinical laboratory, results such as the ones observed in Figures 1 and 2 would have been reported as “small monoclonal IgG κ protein in the γ fraction” with the suggestion to
“repeat testing in 6–12 months if clinically indicated”. The concern is that detection of a therapeutic mAb on SPEP or IFE could trigger additional follow-up testing and concern for a monoclonal gammopathy. However, it is important to note that these abnormalities were only noted for two of the mAbs tested, and only at peak concentrations. In reality, only when concentrations of the biologics in serum are close to 300 μg/mL may they be misidentified for a plasma cell dyscrasia, based on SPEP/IFE results. That will seldom happen for the mAbs tested. It may occur when the blood sample is drawn at peak, hours or a couple of days after the intravenous or subcutaneous infusions. In one study, for patients on INF, the highest concentration in circulation measured 48–72 h after infusion was 172 μg/mL [5]. This concentration of the mAb is
Brought to you by | Washington University in St. Louis Authenticated Download Date | 2/24/16 12:49 PM
8 Willrich et al.: Monoclonal antibody therapeutics as interferences on SPEP and IFE
B
2145.786 Da
2173.889 Da
Intensity, cps
A
vedo 300 µg/mL
IgG κ M-spike (10 g/L)+ vedo 300 µg/mL
SPEP
G
A
M
K
L 2110 2115 2120 2125 2130 2135 2140 2145 2150 2155 2160 2165 2170 2175 2180
Mass/charge, Da
D
2110.538 Da
2095.087 Da
IgG κ M-spike (18 g/L)+ ritu 400 µg/mL
C Intensity, cps
ritu 400 µg/mL
SPEP
G
A
M
K
L 2090
2100
2110
2120
2130
2140
2150
2160
Mass/charge, Da
Figure 7: Immunofixation and miRAMM results from serum samples with endogenous M-spikes and therapeutic mAbs. (A) Immunofixation of IgG κ M-spike measured as 10 g/L by SPEP. Vedolizumab at 300 μg/mL was spiked to this sample, and its unique band is not visible on IFE. (B) Rituximab at 400 μg/mL was added to a sample with endogenous IgG κ gammopathy, M-spike of 18 g/L and may be observed as a small band towards the cathode. (C and D) miRAMM extracted ion chromatograms (+11 charge state). Although the endogenous IgG κ M-spikes peak intensities are more intense than the therapeutic mAbs, the accurate mass is d istinguishable from the endogenous M-protein (blue traces). The pink traces are overlaid negative SPEP/IFE samples spiked with vedolizumab 300 μg/mL (C) or rituximab 400 mcg/mL (D) to facilitate comparisons.
likely not high enough to be seen on IFE, although it is almost certainly identifiable using miRAMM (Figure 3B). The concentration of 500 μg/mL of INF used in one of the experiments is not observed in vivo. The non-visualization of bands on SPEP and IFE for INF was also reported by recent study that evaluated several widely prescribed mAbs in concentrations representing the average therapeutic blood level during a dosing interval and included INF. The study also reported abnormalities on SPEP and IFE for RITU measured at 400 μg/mL [12]. For cancer therapeutic mAbs which are often administered in doses higher than 300 μg/mL (not studied in this report), the identification of a small band on IFE may trigger concern of clinicians.
Overall, biologics are not likely to appear as a monoclonal protein except when SPEP and IFE are performed within a couple of days from infusion (peak), specifically for VEDO and RITU, or perhaps other mAbs administered in high doses. When a monoclonal protein is detected by SPEP or IFE in a patient receiving a therapeutic mAb, it should not be assumed that the abnormality observed is caused by the drug. The initial characterization of the migration patterns on SPEP and IFE of these mAbs may help triage potentially confounding cases. However, this requires some knowledge regarding what position the therapeutic mAb migrates to on the gel in question. Unexpected or multiple bands may occur due to multiple forms of the clonal protein secreted by the plasma cell clone,
Brought to you by | Washington University in St. Louis Authenticated Download Date | 2/24/16 12:49 PM
Willrich et al.: Monoclonal antibody therapeutics as interferences on SPEP and IFE 9
from proteins secreted by a mutated daughter clone, from a completely new plasma cell clone, or (as in this study) by the administration of a therapeutic monoclonal antibody. Hypogammaglobulinemia will cause a lower background from endogenous immunoglobulins, and a therapeutic mAb may evidence as a sharp band characterized on IFE. Conversely, hypergammaglobulinemia may make it more difficult to identify a therapeutic mAb on IFE. However, if a mAb is not mis-identified for a monoclonal gammopathy, the finding will not trigger further testing. If it is, the use of miRAMM can solve ambiguous cases by precisely identifying the therapeutic mAb in a polyclonal background, or distinct from an endogenous monoclonal gammopathy using the accurate molecular mass of the mAb calculated using the amino acid sequences of the light chain. Although theoretically light chains from an endogenous monoclonal protein and a therapeutic mAb can elute at the same time during chromatographic separation, it is very unlikely that the two light chains will also have the same accurate mass. If they differ by only one amino acid, on average that represents a 15 Da mass shift. miRAMM mass accuracy is about ±1 Da [10]. If the accurate mass is indeed the same, which should be a much rarer finding than the overlap with an M-protein on IFE, top-down proteomics can potentially be an added step to miRAMM to discriminate these cases, as the light chains would not have the same signature fragmentation profiles. A library containing the molecular masses of the light chains from commonly administered therapeutic mAbs should prove valuable for reference in the clinical laboratory performing miRAMM experiments. This may be a useful methodology to be applied in the laboratory to help distinguish questionable cases as more and more therapeutic mAbs enter the market. The use of miRAMM can of course help clarify the significance of unexpected electrophoretic findings no matter what their cause. However, it is important to note that the conclusions from this data are only applicable to the five therapeutic mAbs studied. Each mAb used as a therapeutic should be assessed as a potential interference on SPEP and IFE, taking into account migration patterns and peak circulating concentrations. If a therapeutic mAb is identified as a significant concern for SPEP or IFE interference, miRAMM will be valuable for distinguishing the drug from endogenous monoclonal proteins. Author contributions: All the authors have accepted responsibility for the entire content of this submitted manuscript and approved submission.
Research funding: None declared. Employment or leadership: None declared. Honorarium: None declared. Competing interests: The funding organization(s) played no role in the study design; in the collection, analysis, and interpretation of data; in the writing of the report; or in the decision to submit the report for publication.
References 1. Chan AC, Carter PJ. Therapeutic antibodies for autoimmunity and inflammation. Nat Rev Immunol 2010;10:301–16. 2. Wang W, Wang EQ, Balthasar JP. Monoclonal antibody pharmacokinetics and pharmacodynamics. Clin Pharmacol Therap 2008;84:548–58. 3. McCudden CR, Voorhees PM, Hainsworth SA, Whinna HC, Chapman JF, Hammett-Stabler CA, et al. Interference of monoclonal antibody therapies with serum protein electrophoresis tests. Clin Chem 2010;56:1897–9. 4. Genzen JR, Kawaguchi KR, Furman RR. Detection of a monoclonal antibody therapy (ofatumumab) by serum protein and immunofixation electrophoresis. Br J Haematol 2011;155:123–5. 5. Willrich MA, Murray DL, Barnidge DR, Ladwig PM, Snyder MR. Quantitation of infliximab using clonotypic peptides and selective reaction monitoring by LC-MS/MS. Int Immunopharmacol 2015;28:513–20. 6. Katrangi W, Ladwig PM, Barnidge DR, Mills JR, Dasari S, Murray DL, et al. A-260 vedolizumab quantitation in serum using SRM and micro LC-ESI-Q-TOF mass spectrometry. Clin Chem 2015;61:10S. 7. Pittock SJ, Lennon VA, McKeon A, Mandrekar J, Weinshenker BG, Lucchinetti CF, et al. Eculizumab in AQP4-IgG-positive relapsing neuromyelitis optica spectrum disorders: an open-label pilot study. Lancet Neurol 2013;12:554–62. 8. Katzmann JA, Kyle RA, Benson J, Larson DR, Snyder MR, Lust JA, et al. Screening panels for detection of monoclonal gammopathies. Clin Chem 2009;55:1517–22. 9. Barnidge DR, Dasari S, Botz CM, Murray DH, Snyder MR, Katzmann JA, et al. Using mass spectrometry to monitor monoclonal immunoglobulins in patients with a monoclonal gammopathy. J Proteome Res 2014;13:1419–27. 10. Mills JR, Barnidge DR, Murray DL. Detecting monoclonal immunoglobulins in human serum using mass spectrometry. Methods 2015;81:56–65. 11. Barnidge DR, Dasari S, Ramirez-Alvarado M, Fontan A, Willrich MA, Tschumper RC, et al. Phenotyping polyclonal kappa and lambda light chain molecular mass distributions in patient serum using mass spectrometry. J Proteome Res 2014;13: 5198–205. 12. Ruinemans-Koerts J, Verkroost C, Schmidt-Hieltjes Y, Wiegers C, Curvers J, Thelen M, et al. Interference of therapeutic monoclonal immunoglobulins in the investigation of M-proteins. Clin Chem Lab Med 2014;52:e235–7.
Brought to you by | Washington University in St. Louis Authenticated Download Date | 2/24/16 12:49 PM
Maria Alice V. Willrich*, Paula M. Ladwig, Bruna D. Andreguetto, David R. Barnidge, David L. Murray, Jerry A. Katzmann and Melissa R. Snyder
Monoclonal antibody therapeutics as potential interferences on protein electrophoresis and immunofixation DOI 10.1515/cclm-2015-1023 Received October 19, 2015; accepted December 13, 2015
Abstract Background: The use of therapeutic recombinant monoclonal antibodies (mAbs) has triggered concerns of misdiagnosis of a plasma cell dyscrasia in treated patients. The purpose of this study is to determine if infliximab (INF), adalimumab (ADA), eculizumab (ECU), vedolizumab (VEDO), and rituximab (RITU) are detected as monoclonal proteins by serum protein electrophoresis (SPEP) and immunofixation electrophoresis (IFE). Methods: Pooled normal sera were spiked with various concentrations (ranging from trough to peak) of INF, ADA, ECU, VEDO and RITU. The peak concentration for VEDO and RITU was also added to samples with known monoclonal gammopathies. All samples were analyzed by SPEP (Helena Laboratories) and IFE (Sebia); sera containing peak concentrations of mAbs were reflexed to electrospray-time-of-flight mass spectrometry (AbSciex Triple TOF 5600) for the intact light chain monoclonal immunoglobulin rapid accurate mass measurement (miRAMM). Results: For all mAbs tested, no quantifiable M-spikes were observed by SPEP at any concentration analyzed. Small γ fraction abnormalities were noted on SPEP for VEDO at 300 μg/mL and RITU at 400 μg/mL, with identification of small IgG κ proteins on IFE. Using miRAMM for peak samples, therapeutic mAbs light chain accurate masses
*Corresponding author: Maria Alice V. Willrich, PhD, Division of Clinical Biochemistry and Immunology, Department of Laboratory Medicine and Pathology, Mayo Clinic, 200 1st St SW, Rochester, MN 55905, USA, Phone: +507-266-4909, Fax: +507-266-4088, E-mail: [email protected] Paula M. Ladwig, David R. Barnidge, David L. Murray, Jerry A. Katzmann and Melissa R. Snyder: Division of Clinical Biochemistry and Immunology, Department of Laboratory Medicine and Pathology, Mayo Clinic, Rochester, MN, USA Bruna D. Andreguetto: From the Department of Clinical Pathology at Universidade de Campinas, Campinas, SP, Brazil
were identified above the polyclonal background and were distinct from endogenous monoclonal gammopathies. Conclusions: MAbs should not be easily confounded with plasma cell dyscrasias in patients undergoing therapy except when a SPEP and IFE are performed within a couple of days from infusion (peak). In ambiguous cases the use of the miRAMM technology could precisely identify the therapeutic mAb distinct from any endogenous monoclonal protein. Keywords: immunofixation; monoclonal antibody therapies; monoclonal immunoglobulin Rapid Accurate Mass Measurement (miRAMM); protein electrophoresis.
Introduction Therapeutic monoclonal antibodies (mAbs) have rapidly become a clinically important drug class, as the efficacy of their use for treatment of a wide variety of clinical indications becomes proven. There are applications that include autoimmunity and inflammation, cancer, organ transplantation, cardiovascular disease, infectious diseases and ophthalmological diseases [1]. Human immunoglobulins (Ig) are composed of two identical heavy chains and two identical light chains, held together by disulfide bonds. Although there are five types of immunoglobulins defined by their heavy chains (IgG, IgA, IgM, IgD, and IgE) and two types of light chains (κ and λ), the vast majority of the therapeutic mAbs currently on the market are of the IgG κ isotype. The IgG family of antibodies may be further divided, based on the structure of their heavy chains, into four subclasses: IgG1, IgG2, IgG3, and IgG4. Structural differences among IgG heavy chains lead to differences in binding of the subclasses to Fc receptors and, consequently, to subclass- specific differences in processes mediated by Fc receptors [2]. Although most therapeutic mAbs are IgG1, there are many IgG2 and IgG4 currently approved as well. The initial therapeutic mAbs were generated from mouse and rat hybridomas. These first-generation
Brought to you by | Washington University in St. Louis Authenticated Download Date | 2/24/16 12:49 PM
TNF-α, Tumor necrosis factor-α; I.V., intravenous; S.C., subcutaneous; IBD, inflammatory bowel disease; RA, rheumatoid arthritis; aHUS, atypical hemolytic uremic syndrome; PNH, p aroxysmal nocturnal hemoglobinuria; NHL, non-Hodgkin’s lymphoma; CLL, chronic lymphocytic leukemia; GPA, granulomatosis with polyangiitis; MPA; microscopic polyangiitis. aHalf-life obtained from drug package insert or prescribing information.
IBD, RA IBD, RA aHUS, PNH IBD NHL, CLL, GPA, MPA 5–10 mg/kg every 8 weeks 40–80 mg every other week 300–1200 mg every other week 300 mg every 8 weeks 500 mg per adult I.V. S.C. I.V. I.V. I.V. 8–10 days 14 days 11.3 days 25 days 22 days Chimeric IgG1 κ Human IgG1 κ Humanized IgG2/IgG4 κ Humanized IgG1 κ Chimeric IgG1 κ TNF-α TNF-α Complement factor C5 α4β7 integrin CD 20
Administration dose Administration route Average Half-lifea Structure Target molecule
Brought to you by | Washington University in St. Louis Authenticated Download Date | 2/24/16 12:49 PM
Infliximab Adalimumab Eculizumab Vedolizumab Rituximab
INF, ADA, VEDO and RITU were purchased from the Mayo Clinic pharmacy and reconstituted per manufacturer instructions. ECU was supplied in a ready for use formulation. Reagents for SPEP were purchased from Helena Laboratories (Beaumont, TX, USA) and for IFE from Sebia
Materials and methods
Table 1: Characteristics of the studied monoclonal antibodies.
antibodies, named -omab, found only limited success in the clinic because of their short half-lives and high immunogenicity. A number of approaches have been developed to humanize rodent antibodies, from development of chimeric antibodies (–ximab), where constant and variable regions are derived from human and rodent IgG, respectively, to humanized antibodies, where 90%–95% of the antibody is composed of sequence derived from human IgG with only the complementarity determining regions (CDR) being of murine source (–zumab), to fully human antibodies (–umab). Serum protein electrophoresis (SPEP) and immunofixation (IFE) are an important part of the diagnostic work-up for patients with suspected plasma cell dyscrasias. However, SPEP/IFE may also be ordered for a number of other indications, including autoimmune/infectious diseases (polyclonal hypergammaglobulinemia), renal dysfunction (nephrotic proteinuria), and primary immunodeficiency (hypogammaglobulinemia). For many of these diseases, treatment with a therapeutic mAb is becoming common. This raises the question of whether a therapeutic mAb could be mis-interpreted as an endogenous M-protein on SPEP or IFE, potentially triggering unnecessary additional testing and clinical evaluation. A few reports have described the detection of mAb therapeutics on SPEP or IFE in patients under treatment, such as the detection of siltuximab, against interleukin-6 [3] and ofatumumab, an anti-CD20 antibody [4]. Siltuximab is a chimeric (30% murine/70% human) IgG κ mAb. At the time of the reported interference on SPEP, siltuximab was being used in clinical trials for the treatment of refractory multiple myeloma in combination with dexamethasone and was identified in a patient with history of IgD κ myeloma. The purpose of this study was to determine if mAbs used for chronic inflammatory conditions [INF, adalimumab (ADA), eculizumab (ECU), vedolizumab (VEDO), and rituximab (RITU)] (Table 1) could be detected on SPEP or IFE at trough, mid-dose and peak therapeutic concentrations. Secondly the use of time-of-flight (TOF) mass spectrometry was investigated as a way to accurately distinguish between therapeutic and endogenous monoclonal proteins.
Conditions where it is commonly prescribed
2 Willrich et al.: Monoclonal antibody therapeutics as interferences on SPEP and IFE
Willrich et al.: Monoclonal antibody therapeutics as interferences on SPEP and IFE 3 (Norcross, GA, USA). Reagents for mass spectrometry included ammonium bicarbonate, dithiothreitol and formic acid (Sigma-Aldrich Inc., Saint Louis, MO, USA). Melon Gel was purchased from Thermo Fisher (Waltham, MA, USA) and water, 2-propanol and acetonitrile were from Honeywell Burdick and Jackson (Muskegon, MI, USA). Residual normal sera with no abnormalities on SPEP and IFE were pooled and spiked with three different concentrations for each mAb, ranging from trough to peak, as described in Table 2. Concentrations were defined based on studies with in-house serial sample collections from patients undergoing therapy with these mAbs. In order to verify the experiments with normal serum spiked with mAbs, residual serum samples from patients undergoing therapy with INF, ADA, ECU and VEDO were used (n = 4). Concentrations of the mAbs were measured using tryptic peptide mass spectrometry methods (INF and VEDO) [5, 6], a commercial immunoassay for ADA (Esoterix Laboratories, Calabasas, CA, USA) and ELISA for ECU [7]. All samples were analyzed using standard operating procedures from the clinical laboratory for SPEP (Helena Laboratories) and IFE (Sebia) [8]. A non-quantifiable SPEP abnormality, denoted by term “small”, is defined by the laboratory as an observable electrophoretic restriction or asymmetry that is too small to fractionate as an M-spike. A third group of de-identified residual waste serum samples with small γ migrating abnormalities and subsequent monoclonal IgG κ’s identified on IFE (n = 10) was collected and used for comparison to a fourth group of residual samples without SPEP abnormalities (n = 10) spiked with 500 μg/mL of INF. Finally, two samples with known IgG κ monoclonal gammopathies and quantifiable M-spikes were spiked with peak concentrations of VEDO (300 μg/mL) or RITU (400 μg/mL). Serum samples were analyzed using electrospray-time-of-flight mass spectrometry (AbSciex Triple TOF 5600) for the intact light chain using the monoclonal immunoglobulin Rapid Accurate Mass Measurement (miRAMM) method [9]. Briefly, this method utilizes microflow liquid chromatography electrospray-time-of-flight mass spectrometry to identify the accurate molecular mass of the light chain portion of a monoclonal immunoglobulin. Sample preparation included enrichment of IgG from serum samples using Melon Gel resin (Thermo Fisher). 20 μL of serum was added to 200 μL of resin and mixed for 5 min on an orbital shaker. 20 μL of the supernatant was added to 20 μL of 50 mM ammonium bicarbonate and 10 μL of 200 mM dithiothreitol (DTT) to release light chains from the heavy chains. The mixture was incubated for 15 min at 55 °C before injection onto an Eksigent Ekspert 200 microLC system (Redwood City, CA, USA) in a Poroshell 300SB-C3 analytical column (Agilent) kept in a column heater at 60 °C, using a flow rate of 25 μL/min in a total chromatography run time of 24 min. Mobile phases A (aqueous) and B (90% acetonitrile/10% 2-propanol/0.1% formic acid) were used in a gradient for light chain elution as previously described [10].
The ABSciex tripleTOF 5600 (ABSciex, Framingham, MA, USA) was operated in ESI positive mode with a Turbo V dual-ion source. Source conditions were ISVF 5500/temperature 50 °C/CUR 45/GS1 35/GS2 30. TOF MS scans ranged from m/z of 600–2500 with 200 ms of acquisition time. Analyst TF v1.6 software was used for instrument control and data viewing. The +11 charge state was used for identification of the mAbs in the total ion chromatograms, with the spectra being shown as mass-to-charge (m/z) ratios. The +11 charge state was chosen as it consistently showed a total κ/λ ratio in close agreement to those found by immunonephelometry [11]. The mass spectra of the multiply charged ions were deconvoluted using the Analyst software to a molecular mass characteristic of each light chain in the individual’s immunoglobulin repertoire or the therapeutic mAb. The mass of the intact light chains for each mAbs used as therapeutics were characterized using the pure pharmaceutical preparations (Table 2) diluted in 50 mM ammonium bicarbonate.
Results Pooled normal sera spiked with peak, mid-dose, and trough concentrations for INF, ADA, ECU, VEDO, or RITU were analyzed by routine SPEP and IFE. On SPEP, no quantifiable M-spikes at any concentration of the five therapeutic mAbs were observed. In addition, no small abnormalities were detected on SPEP for any of the “expected” trough or mid-dose concentrations of the therapeutic mAbs. The only small abnormalities noted were for VEDO and RITU at the peak concentrations of 300 μg/mL and 400 μg/mL, respectively (Figure 1). The small band caused by VEDO migrates in the center of the γ fraction and the band from RITU is very cathodal, migrating as a blunt end on the electrophoretic densitogram. On IFE, these small electrophoretic abnormalities seen with VEDO and RITU at the peak concentrations were identified as small IgG κ bands (Figure 2). No abnormalities were identified on IFE for any of the mAbs at trough or mid-dose concentrations (data not shown). A normal serum sample and the samples spiked with peak concentrations of the five therapeutic mAbs were then subjected to miRAMM. The normal serum shows two distinct Gaussian peaks corresponding to the λ and κ polyclonal immunoglobulin repertoires as previously described (Figure 3A) [11]. In the spiked samples, the mAbs used as
Table 2: Typical concentrations observed for monoclonal antibody therapeutics and intact light chain accurate mass measured by miRAMM.
Target therapeutic thresholds at trough
Typical concentrations at mid-term
Typical concentrations at peak
Deconvoluted accurate mass of intact light chains on TOF (+11 charge state m/z)
Infliximab Adalimumab Eculizumab Vedolizumab Rituximab
5 μg/mL 10 μg/mL 50 μg/mL 5 μg/mL 10 μg/mL
30 μg/mL 30 μg/mL 100 μg/mL 30 μg/mL 100 μg/mL
100 μg/mL 100 μg/mL 200 μg/mL 300 μg/mL 400 μg/mL
23,434 Da (2131.4) 23,406 Da (2129.0) 23,130 Da (2103.8) 23,901 Da (2173.9) 23,035 Da (2095.1)
Brought to you by | Washington University in St. Louis Authenticated Download Date | 2/24/16 12:49 PM
4 Willrich et al.: Monoclonal antibody therapeutics as interferences on SPEP and IFE
SPEP 5
30
100
10
inf
30
100
50
100 200
ada
5
ecu
30
300
10
100 400
vedo
ritu
Figure 1: SPEP gels of originally negative SPEP/IFE samples after spiking of mAbs in concentrations expected to be found at trough, middose and peak of therapy regimens. Concentrations shown in μg/mL.
A
B
inf 100
SPEP
G
A
M
ada 100
K
L
SPEP
C
G
A
M
K
L
ecu 200
(µg/mL)
SPEP
D
G
A
M
K
L
E vedo 300
SPEP
G
A
M
ritu 400
K
L
SPEP
G
A
M
K
L
Figure 2: Immunofixation gels. (A) Infliximab at 100 μg/mL was spiked in serum. (B) Adalimumab at 100 μg/mL in serum. (C) Eculizumab at 200 μg/mL. (D) Vedolizumab at 300 μg/mL. (E) Rituximab at 400 μg/mL. Small abnormalities in the γ fraction are noted for vedolizumab and rituximab.
therapies are detectable as a peak above the polyclonal background of endogenous immunoglobulins. Each mAb, with its unique κ light chain, demonstrates a different m/z ratio (Figure 3B–F). The m/z of the +11 charge state as measured by TOF mass spectrometry for each therapeutic mAb in serum was within ±0.5 Da of the mass as determined by measuring the therapeutic mAbs in buffer (Table 2). Samples from patients undergoing therapy with the studied mAbs are shown in Figure 4. SPEP and IFE findings show no abnormalities for INF or ADA. ECU was present in circulation in patient’s serum at 249 μg/mL,
and was additionally observed as a faint small IgG κ on IFE. VEDO was measured at 131 μg/mL, and although not a sharp band on IFE, a dense IgG κ staining is observed. MiRAMM results are presented for the samples and all mAbs accurate mass was identifiable above the polyclonal background, even if not observed on IFE (Figure 4). A sample from a patient undergoing RITU therapy was not available. Next, the ability of miRAMM to distinguish between therapeutic and endogenous mAbs was assessed. Previously characterized serum samples with confirmed small
Brought to you by | Washington University in St. Louis Authenticated Download Date | 2/24/16 12:49 PM
Willrich et al.: Monoclonal antibody therapeutics as interferences on SPEP and IFE 5
A
Normal SPEP/IFE
λ
ada 100 µg/mL 2128.9 Da
vedo 300 µg/mL 2173.9 Da
B
inf 100 µg/mL 2131.4 Da
κ
C
D
ecu 200 µg/mL 2103.7 Da
E
ritu 400 µg/mL 2095.1 Da
F
Figure 3: MiRAMM deconvoluted extracted ion chromatograms. (A) Negative SPEP/IFE serum light chains miRAMM extracted ion chromatogram. The Gaussian distribution of λ and κ light chains is shown. (B) Infliximab at 100 μg/mL in serum shows up as a monoclonal peak above the polyclonal background. (C) Adalimumab at 100 μg/mL. (D) Eculizumab at 200 μg/mL. (E) Vedolizumab at 300 μg/mL. (F) Rituximab at 400 μg/mL. X axis of figures, mass-to-charge ratios (Da). Y axis, intensity (cps).
IgG κ abnormalities migrating in the γ f raction are shown on a representative IFE gel in Figure 5A. Figure 5B, on the other hand, shows IFE results of normal sera spiked with 500 μg/mL of INF. Most of the samples spiked with INF demonstrate a detectable IgG κ monoclonal protein, with the exception of samples #4 and #5, in which an endogenous polyclonal hypergammaglobulinemia (γ fraction of 29 g/L and 15 g/L, respectively) masks the band. The spiked group of samples most likely would have been reported as small IgG κ abnormalities in the γ fraction if these were clinical specimens. The endogenous and spiked samples were then analyzed by miRAMM. The results from the 10 samples with endogenous small IgG κ are shown as overlaid extracted ion chromatograms (+11 charge state). Each sample shows a sharp peak which represents the light chain from the monoclonal IgG κ. However, each light chain has a unique mass, representing diversity across the different monoclonal proteins (Figure 6A). In contrast, for the samples spiked with 500 μg/mL of INF, all of the miRAMM chromatograms show a single peak above the
polyclonal background, corresponding to the INF mass at the +11 charge state. When the spectra are overlaid, all samples spiked with INF show the same mass for the monoclonal κ light chain peak (Figure 6B). The mAb is observed even in the presence of hypergammaglobulinemia, noticeable especially on the red trace chromatogram. Lastly, the ability of miRAMM to identify a therapeutic mAb in the presence of a large endogenous monoclonal protein with a quantifiable M-spike was addressed. M-spikes characterized as IgG κ isotypes were spiked with VEDO at 300 μg/mL or RITU at 400 μg/mL. IFE and miRAMM are shown in Figure 7. The sample spiked with VEDO did not have an additional visible band on IFE. The mAb co-migrates with the M-spike on IFE, however miRAMM was able to separate the endogenous M-spike from the therapeutic mAb based on each clone’s unique mass (Figure 7A and B). The sample spiked with RITU, instead, had an additional band in the γ fraction on IFE. MiRAMM confirmed the additional band was from RITU based on its light chain mass (Figure 7C and D).
Brought to you by | Washington University in St. Louis Authenticated Download Date | 2/24/16 12:49 PM
6 Willrich et al.: Monoclonal antibody therapeutics as interferences on SPEP and IFE
2131.4 Da Patient A Inf 125 µg/mL
2128.8 Da Patient B Ada 36 µg/mL
2103.7 Da Patient C Ecu 249 µg/mL
2173.9 Da Patient D Vedo 131 µg/mL
SPEP G
A
M
K
L
Figure 4: Immunofixation and miRAMM results from residual serum of patients undergoing therapy with different mAbs. Samples were collected at different time-points after therapy, and reflect concentrations usually found at peak (hours-days after dose). Concentrations of the mAbs were measured using tryptic methods and LC-MS/MS (infliximab and vedolizumab), a commercial electrochemiluminescent assay (adalimumab) and ELISA (eculizumab).
A
B
Figure 5: Immunofixation gels. (A) Lanes 1–8, serum samples reported as “small IgG κ in the γ fraction”. Lane 9, IgM κ positive control for immunofixation. (B) Lanes 1–8, serum samples originally negative by immunofixation after spiking of 500 μg/mL of infliximab. On lane 4 it is possible to note hypergammaglobulinemia (original γ fraction of 29 g/L). Lane 9, IgM κ positive control. Note: 10 samples were selected for each group. Only 8 IFE reports shown due to space constrains. Results of omitted samples (9–10) are similar to the reports shown. Brought to you by | Washington University in St. Louis Authenticated Download Date | 2/24/16 12:49 PM
Willrich et al.: Monoclonal antibody therapeutics as interferences on SPEP and IFE 7
A
170
7
160 150 140 130
Intensity, cps
120 110 100
4
3
5
90 80
2
10
70 60 50 40
9a 9b
30 20
6 9c 1 8a
8b
1
2154.906 Da
2
2202.715 Da
3
2174.734 Da
4
2114.526 Da
5
2131.976 Da
6
2150.935 Da
7
2127.539 Da
8
(a) 2128.340 Da (b) 2136.976 Da
9
(a) 2091.937 Da (b) 2097.042 Da (c) 2147.102 Da
10
2130.643 Da
1
2131.411 Da
2
2131.419 Da
3
2131.437 Da
4
2131.451 Da
5
2131.445 Da
6
2131.455 Da
7
2131.451 Da
8
2131.422 Da
9
2131.425 Da
10
2131.428 Da
10 0 2040 2050 2060 2070 2080 2090 2100 2110 2120 2130 2140 2150 2160 2170 2180 2190 2200 2210 2220
Mass/charge, Da
Intensity, cps
B
35 34 33 32 31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
1–10
2020 2030 2040 2050 2060 2070 2080 2090 2100 2110 2120 2130 2140 2150 2160 2170 2180 2190 2200
Mass/charge, Da
Figure 6: MiRAMM results from residual serum samples. (A) miRAMM extracted ion chromatogram showing the +11 charge state zoomed spectra of ten samples reported as small IgG κ in the γ fraction. The table on the right shows the masses of the monoclonal peaks observed above the polyclonal background for each of the samples. Sample 8 had 2 unique clones; sample 9 had 3 clones. (B) miRAMM (+11 charge state) of samples spiked with 500 μg/mL of infliximab. The table on the right shows the measured mass for the monoclonal peaks for all 10 samples.
Discussion Therapeutic mAbs are gaining market share because of proven efficacy in a wide range of clinical disorders. These applications are not only in oncology but also in autoimmune conditions such as rheumatoid arthritis and inflammatory bowel disease. Some of the mAbs are administered at very high doses, especially in the treatment of tumors, usually for a short period of time, as adjuncts to chemo- or radiation therapies. Others, like the ones presented in this study, are used long-term, for treatment of chronic diseases, and are administered at lower doses. In our clinical laboratory, results such as the ones observed in Figures 1 and 2 would have been reported as “small monoclonal IgG κ protein in the γ fraction” with the suggestion to
“repeat testing in 6–12 months if clinically indicated”. The concern is that detection of a therapeutic mAb on SPEP or IFE could trigger additional follow-up testing and concern for a monoclonal gammopathy. However, it is important to note that these abnormalities were only noted for two of the mAbs tested, and only at peak concentrations. In reality, only when concentrations of the biologics in serum are close to 300 μg/mL may they be misidentified for a plasma cell dyscrasia, based on SPEP/IFE results. That will seldom happen for the mAbs tested. It may occur when the blood sample is drawn at peak, hours or a couple of days after the intravenous or subcutaneous infusions. In one study, for patients on INF, the highest concentration in circulation measured 48–72 h after infusion was 172 μg/mL [5]. This concentration of the mAb is
Brought to you by | Washington University in St. Louis Authenticated Download Date | 2/24/16 12:49 PM
8 Willrich et al.: Monoclonal antibody therapeutics as interferences on SPEP and IFE
B
2145.786 Da
2173.889 Da
Intensity, cps
A
vedo 300 µg/mL
IgG κ M-spike (10 g/L)+ vedo 300 µg/mL
SPEP
G
A
M
K
L 2110 2115 2120 2125 2130 2135 2140 2145 2150 2155 2160 2165 2170 2175 2180
Mass/charge, Da
D
2110.538 Da
2095.087 Da
IgG κ M-spike (18 g/L)+ ritu 400 µg/mL
C Intensity, cps
ritu 400 µg/mL
SPEP
G
A
M
K
L 2090
2100
2110
2120
2130
2140
2150
2160
Mass/charge, Da
Figure 7: Immunofixation and miRAMM results from serum samples with endogenous M-spikes and therapeutic mAbs. (A) Immunofixation of IgG κ M-spike measured as 10 g/L by SPEP. Vedolizumab at 300 μg/mL was spiked to this sample, and its unique band is not visible on IFE. (B) Rituximab at 400 μg/mL was added to a sample with endogenous IgG κ gammopathy, M-spike of 18 g/L and may be observed as a small band towards the cathode. (C and D) miRAMM extracted ion chromatograms (+11 charge state). Although the endogenous IgG κ M-spikes peak intensities are more intense than the therapeutic mAbs, the accurate mass is d istinguishable from the endogenous M-protein (blue traces). The pink traces are overlaid negative SPEP/IFE samples spiked with vedolizumab 300 μg/mL (C) or rituximab 400 mcg/mL (D) to facilitate comparisons.
likely not high enough to be seen on IFE, although it is almost certainly identifiable using miRAMM (Figure 3B). The concentration of 500 μg/mL of INF used in one of the experiments is not observed in vivo. The non-visualization of bands on SPEP and IFE for INF was also reported by recent study that evaluated several widely prescribed mAbs in concentrations representing the average therapeutic blood level during a dosing interval and included INF. The study also reported abnormalities on SPEP and IFE for RITU measured at 400 μg/mL [12]. For cancer therapeutic mAbs which are often administered in doses higher than 300 μg/mL (not studied in this report), the identification of a small band on IFE may trigger concern of clinicians.
Overall, biologics are not likely to appear as a monoclonal protein except when SPEP and IFE are performed within a couple of days from infusion (peak), specifically for VEDO and RITU, or perhaps other mAbs administered in high doses. When a monoclonal protein is detected by SPEP or IFE in a patient receiving a therapeutic mAb, it should not be assumed that the abnormality observed is caused by the drug. The initial characterization of the migration patterns on SPEP and IFE of these mAbs may help triage potentially confounding cases. However, this requires some knowledge regarding what position the therapeutic mAb migrates to on the gel in question. Unexpected or multiple bands may occur due to multiple forms of the clonal protein secreted by the plasma cell clone,
Brought to you by | Washington University in St. Louis Authenticated Download Date | 2/24/16 12:49 PM
Willrich et al.: Monoclonal antibody therapeutics as interferences on SPEP and IFE 9
from proteins secreted by a mutated daughter clone, from a completely new plasma cell clone, or (as in this study) by the administration of a therapeutic monoclonal antibody. Hypogammaglobulinemia will cause a lower background from endogenous immunoglobulins, and a therapeutic mAb may evidence as a sharp band characterized on IFE. Conversely, hypergammaglobulinemia may make it more difficult to identify a therapeutic mAb on IFE. However, if a mAb is not mis-identified for a monoclonal gammopathy, the finding will not trigger further testing. If it is, the use of miRAMM can solve ambiguous cases by precisely identifying the therapeutic mAb in a polyclonal background, or distinct from an endogenous monoclonal gammopathy using the accurate molecular mass of the mAb calculated using the amino acid sequences of the light chain. Although theoretically light chains from an endogenous monoclonal protein and a therapeutic mAb can elute at the same time during chromatographic separation, it is very unlikely that the two light chains will also have the same accurate mass. If they differ by only one amino acid, on average that represents a 15 Da mass shift. miRAMM mass accuracy is about ±1 Da [10]. If the accurate mass is indeed the same, which should be a much rarer finding than the overlap with an M-protein on IFE, top-down proteomics can potentially be an added step to miRAMM to discriminate these cases, as the light chains would not have the same signature fragmentation profiles. A library containing the molecular masses of the light chains from commonly administered therapeutic mAbs should prove valuable for reference in the clinical laboratory performing miRAMM experiments. This may be a useful methodology to be applied in the laboratory to help distinguish questionable cases as more and more therapeutic mAbs enter the market. The use of miRAMM can of course help clarify the significance of unexpected electrophoretic findings no matter what their cause. However, it is important to note that the conclusions from this data are only applicable to the five therapeutic mAbs studied. Each mAb used as a therapeutic should be assessed as a potential interference on SPEP and IFE, taking into account migration patterns and peak circulating concentrations. If a therapeutic mAb is identified as a significant concern for SPEP or IFE interference, miRAMM will be valuable for distinguishing the drug from endogenous monoclonal proteins. Author contributions: All the authors have accepted responsibility for the entire content of this submitted manuscript and approved submission.
Research funding: None declared. Employment or leadership: None declared. Honorarium: None declared. Competing interests: The funding organization(s) played no role in the study design; in the collection, analysis, and interpretation of data; in the writing of the report; or in the decision to submit the report for publication.
References 1. Chan AC, Carter PJ. Therapeutic antibodies for autoimmunity and inflammation. Nat Rev Immunol 2010;10:301–16. 2. Wang W, Wang EQ, Balthasar JP. Monoclonal antibody pharmacokinetics and pharmacodynamics. Clin Pharmacol Therap 2008;84:548–58. 3. McCudden CR, Voorhees PM, Hainsworth SA, Whinna HC, Chapman JF, Hammett-Stabler CA, et al. Interference of monoclonal antibody therapies with serum protein electrophoresis tests. Clin Chem 2010;56:1897–9. 4. Genzen JR, Kawaguchi KR, Furman RR. Detection of a monoclonal antibody therapy (ofatumumab) by serum protein and immunofixation electrophoresis. Br J Haematol 2011;155:123–5. 5. Willrich MA, Murray DL, Barnidge DR, Ladwig PM, Snyder MR. Quantitation of infliximab using clonotypic peptides and selective reaction monitoring by LC-MS/MS. Int Immunopharmacol 2015;28:513–20. 6. Katrangi W, Ladwig PM, Barnidge DR, Mills JR, Dasari S, Murray DL, et al. A-260 vedolizumab quantitation in serum using SRM and micro LC-ESI-Q-TOF mass spectrometry. Clin Chem 2015;61:10S. 7. Pittock SJ, Lennon VA, McKeon A, Mandrekar J, Weinshenker BG, Lucchinetti CF, et al. Eculizumab in AQP4-IgG-positive relapsing neuromyelitis optica spectrum disorders: an open-label pilot study. Lancet Neurol 2013;12:554–62. 8. Katzmann JA, Kyle RA, Benson J, Larson DR, Snyder MR, Lust JA, et al. Screening panels for detection of monoclonal gammopathies. Clin Chem 2009;55:1517–22. 9. Barnidge DR, Dasari S, Botz CM, Murray DH, Snyder MR, Katzmann JA, et al. Using mass spectrometry to monitor monoclonal immunoglobulins in patients with a monoclonal gammopathy. J Proteome Res 2014;13:1419–27. 10. Mills JR, Barnidge DR, Murray DL. Detecting monoclonal immunoglobulins in human serum using mass spectrometry. Methods 2015;81:56–65. 11. Barnidge DR, Dasari S, Ramirez-Alvarado M, Fontan A, Willrich MA, Tschumper RC, et al. Phenotyping polyclonal kappa and lambda light chain molecular mass distributions in patient serum using mass spectrometry. J Proteome Res 2014;13: 5198–205. 12. Ruinemans-Koerts J, Verkroost C, Schmidt-Hieltjes Y, Wiegers C, Curvers J, Thelen M, et al. Interference of therapeutic monoclonal immunoglobulins in the investigation of M-proteins. Clin Chem Lab Med 2014;52:e235–7.
Brought to you by | Washington University in St. Louis Authenticated Download Date | 2/24/16 12:49 PM
