A Rapid, Simple, Trace, Cost-Effective, and High-Throughput Stable Isotope-Dilution Liquid Chromatography–Tandem Mass Spectrometry Method for Serum Methylmalonic Acid Quantification and Its Clinical Applications

Highlights What are the main findings? An established method for trace, simple, rapid, cheap, sensitive, accurate, robust, and high-throughput for methylmalonic acid quantification. Good chromatographic separation of MMA and its intrinsic isomer and good signals of MMA were achieved using a simple isocratic elution strategy. Materials and reagents that are complex or not always accessible and procedures in previous methods such as derivatization, multistep SPE, incubations, evaporations, drying, or reconstitutions are not required in this MMA quantification method. What is the implication of the main finding? This method is suitable for large-scale MMA testing. Abstract Background: Methylmalonic acid (MMA) is an essential indicator of vitamin B12 (VB12) deficiency and inherited metabolic disorders (IMDs). The increasing number of requests for MMA testing call for higher requirements for convenient MMA testing methods. This study aims to develop a convenient quantification method for serum MMA. Methods: The method was established based on the stable isotope-dilution liquid chromatography–tandem mass spectroscopy (ID-LC-MS/MS) technique. The LC-MS/MS parameters and sample preparation were optimized. Specificity, sensitivity, robustness, accuracy, and clinical applicability were validated according to CLSI C62-A guidelines. MMA levels in VB12-sufficient subjects and VB12-deficient subjects were measured. Results: MMA and its intrinsic isomer, i.e., succinic acid (SA), were completely separated. The average slope, intercept, and correlation relationship (R) with 95% confidence intervals, during the two months, were 0.992 (0.926–1.059), −0.004 (−0.012–0.004), and 0.997 (0.995–0.999), respectively. The limit of detection and quantification were <0.058 μmol/L and 0.085 μmol/L, respectively. Intra-run, inter-run, and total imprecisions were 1.42–2.69%, 3.09–5.27%, and 3.22–5.47%, respectively. The mean spiked recoveries at the three levels were 101.51%, 92.40%, and 105.95%, respectively. The IS-corrected matrix effects were small. The VB12-deficient subjects showed higher MMA levels than VB12-sufficient subjects. Conclusions: A convenient LC-MS/MS method for serum MMA measurement was developed and validated, which could be suitable for large-scale MMA testing and evaluating MMA levels in VB12-deficient patients.


Introduction
Methylmalonic acid (MMA), an abnormal metabolic product of defective cobalamin metabolism and methylmalonate, is considered to be a specific diagnostic biomarker of vitamin B12 (VB12) deficiency [1][2][3][4] and methylmalonic acidemia [5]. VB12 (i.e., cobalamin) is a key cofactor of the enzymatic conversion of methylmalonyl-CoA to succinyl-CoA and the conversion of homocysteine to methionine [6]. An insufficiency of VB12 can lead to elevated methylmalonyl-CoA and homocysteine, causing high levels of MMA and total homocysteine (tHcy) in the blood [3,4]. VB12 deficiency is a common and serious condition [3,4,7] with a high prevalence across many populations, for example, 70-80% in African and Asian children, 40% in Latin American children and adults, and close to 20% in the elderly [3,7]. The measurement of MMA level is also helpful in the screening, confirmation diagnosis, and therapy monitoring of inherited metabolic disorder (IMD) [8].
Early identification of VB12 deficiency is vital for early determination of causes and early prevention and remission of serious presentations; VB12, tHcy, and MMA are all helpful diagnostic indicators for these disorders. The VB12 assay is the most commonly used test. However, previous studies have suggested that MMA is a more sensitive and representative biomarker for VB12 deficiency than VB12 and tHcy [3,4], because: (1) Testing for tHcy is less specific since its concentrations can be easily influenced by many environmental factors [3,9]; (2) VB12 is a less stable indicator and is easily affected by environmental factors [10], while MMA is very stable [11]; (3) MMA is more sensitive than VB12 because MMA can increase before VB12 falls [12,13]. (4) Elevated MMA can persist for several days even after replacement is started [14]. (5) VB12 measurements cannot reflect the true status of VB12. VB12 is measured by automated immunoassays based on competitive-binding immune chemiluminescence [3,15,16]. However, immunoassays lack specificity since they simultaneously measure VB12, as well as holohaptocorrin and holotranscobalamin [3,10]. Under these circumstances, increasingly more requests for MMA testing, and therefore, effective and convenient MMA testing is in demand.
Different from the previous paths of improvement, we took advantage of a simple mobile phase strategy to improve simple MMA detection, which required fewer reagents and a smaller sample. There was no need for complex and dangerous derivation reagents, costly/not always accessible ultrafiltration materials, or processes that are time-consuming and laborious, such as evaporations, incubations, dryings, and reconstitutions. The established method in this study demonstrated several advantages for MMA detection, for example, convenient, environmentally friendly, economical, and more cost-effective for the assessment of VB12 deficiency; only several simple reagents in small volume were needed and sample preparation could be completed in 20 min. Since LC-MS/MS testing is currently mainly manual, such improvements are significant, especially, when there are numerous requests for MMA testing. Table 1. A literature review of previous methylmalonic acid quantification based on the liquid chromatography-tandem mass spectrometry method (from 2000 to  2022). Abbreviations of the names of equipment/columns are not listed and abbreviations of reagents are list in the table for reading convenience. Note: LC-MS/MS, liquid chromatography-tandem mass spectrometry; LC, liquid chromatography; MS, mass spectrometry; MMA, methylmalonic acid; LOD, limit of detection; LOQ, limit of quantification; HCl, hydrochloride; SPE, solid phase extraction; SRM, selected reaction monitoring; CVs, coefficient of variations; NA, not available; MRM, multiple reaction monitoring; DBS, dried blood spot; SIM, single-ion monitoring; HPLC, high-performance liquid chromatography; tHcy, total homocysteine. Little known about the ways to/details of method validation. 8.
Low sensitivity: the highest LOQ and LOD as compared with the other methods, and is not suitable for the investigation of populations with normal (usually <2 µmol/L) to deficient levels of vitamin B12 (varied MMA levels). 9.
Organic solvent in larger volume were required as the mobile phase as compared with our method. Requiring reagents are cheap and easily accessible. 8.
The SIM mode is less reliable than MRM mode. 9.
In particular, using ammonium acetate buffer would require more time and effort for the equipment maintaince/startup before/after use, and would be required to prevent salting out during use, otherwise, it would easily cause contamination or scrapping of the instrument or column! 10.
Organic solvent in large volume were required as the mobile phase (e.g., to prepare 1000 mL mobile phase, 800 mL acetonitrile is required). 11.
The total LC-MS/MS assay time, including column washing and reconditioning, was 10 min, which is longer as compared with other methods. Requiring ultrafilters which are not cheap and not always accessible. 8.
Using ultrafilter requires long time for centrifugation. 9.
Little known about method performance. 10.
The self-made dialyzed plasma used for calibration is hard and complex to prepare. Ammonium formate 4.

2.
A 400 µL aliquot of 0.5 mol/L H3PO4 in MTBE was added, vortexed for 2 min followed by 3 min centrifugation to separate residual water.

3.
A 300 µL aliquot of the supernatant was transferred to tubes and the solvent was evaporated to residue (10-15 min) by centrifugal vacuum evaporation.

4.
Residues were reconstituted in 40 µL of 3 nmol/L HCl in n-butanol, vortexed to mix, and incubated in a 60 • C water bath for 30 min. Solvent was removed by centrifugal vacuum evaporation in an acid resistant system, as above.

5.
Residues were then reconstituted in 100 µL of 100 nmol/L CUDA in mobile phase and vortexed to assist dissolution. 6.
Extracts were transferred to spin filter tubes, spun for 3 min at 4500 rcf, transferred to autosampler vials, capped, and stored at −20 • C until analysis.  [29] TSQ Quantum Access triple quadrupole mass spectrometer(ThermoFisher Scientific) with a transcend TLX-4 multichannel HPLC system (ThermoFisher Scientific) A Cyclone-MAX TurboFlow column (50 × 0.5 mm, ThermoFisher Scientific) was used for online extraction, and a mixing column (Agilent, Santa Clara, CA, USA) was placedbetween the injector and the TurboFlow column Formic acid 7.
Saline solution 1. 500 µL sample (serum or plasma) was mixed with 500 µL of water and 25 µL of working IS solution.
Requiring to prepare many reagents and solutions for multistep SPE. Procedure steps: simple and consists of ultrafiltration and centrifugations but needs incubation.
Requiring Microcon ultrafilters which are not cheap and always not accessible. 9.
More cost for vitamin-investigating patients since tHcy can be measured by economic immunoassays and other biomarkers are not relevant to vitaminB12 assessment.
The mixture was vortexed directly, incubated for 30 min and, meanwhile, vortexed again after every 10 min. 3.
The following deproteinization of plasma was performed by ultrafiltration using Microcon centrifugal filter tubes.

4.
The solution was transferred to the sample reservoir of the Microcon ultrafilter device and the tube was centrifuged at 15 • C for 30 min at 16,300× g.

5.
Then, the filtrate was pipetted into a vial for LC-MS/MS analysis. Procedure steps: consists of protein precipitation, and ultrafiltration, a reduction step has to be carried out to ensure the measurement of tHcy, complex.

6.
Requiring at least 1 h for sample treatment. 7.
More cost for vitamin deficiency-investigating patients since tHcy can be measured by economic immunoassays. 8.
Requiring Microcon ultrafilters which are not cheap and always not accessible. Formic acid 3.
Dialyzed plasma (MMA and SA free).  DTT (50 µL, 500 mmol/L) was added and mixed at room temperature for 15 min to completely reduce disulfides.
After 10 min of centrifugation at 13,000× g, 150 µL of the clear supernatant was transferred and concentrated to dryness under nitrogen (requires at least 20 min), and then reconstituted using 100 µL water containing 1% formic acid. Finally, 5 µL was injected into the system for analysis via LC-MS/MS.

5.
A urine sample was diluted 10 folds using ultrapure water, and then prepared in the same manner as serum samples without protein precipitation and drying. The sample was finally diluted using 500 µL water containing 1% formic acid, and then transferred to a 700 µL 96-well collection plate.  3-Nitrophenylhydrazine (3-NPH)
All the sample deproteinization and derivatization reactions were performed using the fully automated LCMS pretreatment system CLAMTM-2030 (Shimadzu Corp., Kyoto, Japan).
(1) 10 µL of serum sample and 90 µL of methanol containing the internal standard were dispensed into a dedicated vial and stirred for 60 s. (2) Samples were filtered using vacuum pressure for 60 s.  For the elution, 3 mL of NH 4 OH 0.01 mol/L was added to the glass vial, and then agitated for 10 min at 300 rpm on a model G2 gyratory shaker.

5.
The filtrate was retrieved in a 1 mL 96-well collection plate, which was covered with an XP-100 sealing film to prevent clogging of the needle seat by cellulose paper particles during the LC analyses.  Not that cost-effective since requiring creatinine to be measured along with MMA, raising the cost. The solutions were shaken for one minute, and then centrifuged for 10 min at 3000× g.

4.
The supernatant (300 µL) was added to another plate and evaporated with nitrogen gas for 30-45 min.

5.
The residual was reconstituted by 200 µL 0.2% formic acid in water and the plate was shaken for 1 min. Subsequently, the plate was centrifuged for 10 min at 3000× g, prior to LC-MS/MS analysis.  Requiring at least 50 min for sample treatment. 7.
Their claimed requirements on steps: required to ensure that the acetonitrile has evaporated because the presence of this solution causes a double peak in the chromatogram. 8.
As mentioned above in Ma [34], the capability of acetonitrile in protein precipitation is weak, the direct injection of supernatant obtained by simple centrifugation may have potential risk of contaminating the mass spectrum or blocking the column; therefore, acetonitrile should not be selected for protein precipitation unless filter tubes is used for centrifugations. 9.
Due to more time required by evaporation with nitrogen gas, reconstitutions, and limited dryness devices, the method may not be convenient or suitable for large request volumes of MMA testing. Evaporation and reconstitution for several samples seems rapid but will be cumbersome for numerous samples.

Samples
Pooled and individual serum samples for method establishment and individual specimens for clinical application were obtained from fresh leftover specimens in the Department of Laboratory Medicine, Beijing Hospital, Beijing, China. The collection of leftover sera was approved by the Ethics Committee of the Beijing Hospital.

Calibrators, Internal Standard (IS), and Quality Control (QC) Materials Preparation
Working standard solutions of MMA (218.69 ng/g) used for calibration and working standard solutions of MMA-13 C 4 (148.95 ng/g) were prepared gravimetrically in water. All solutions were aliquoted into 2 mL brown ampoules and stored at −80 • C. For each sample batch, the calibrators were freshly prepared. Eight working calibrators were prepared using 100 µL working standard solutions and subsequently diluted with 800, 700, 600, 500, 400, 300, 200, and 100 µL water, respectively.
Standard solutions of MMA (2003.60 ng/g) used as QC additive standard solutions were prepared gravimetrically in water. To make low-, medium-, and high-levels of QC materials, 3 µL, 5 µL, and 8 µL of QC additive solutions were gravimetrically added to three gravimetrically prepared 400 µL serum pools, respectively.

Sample Preparation
Fifty microliters of IS working solution, 50 µL of samples/calibrators, and 300 µL methanol were added to a 2.0 mL Eppendorf tube. The mixture was vortexed for 20 s and centrifuged at 148,000 rpm, at 4 • C, for 15 min. The upper phase of the mixture was poured into a new 2.0 mL Eppendorf tube and centrifuged at 148,000 rpm, at 4 • C, for 5 min. Fifty microliters of the upper phase was used for the LC-MS/MS analysis. The injection volume was 1 µL.

LC-MS/MS Conditions
The LC-MS/MS analysis was performed on a 6500 plus triple quadrupole mass spectrometer (AB Sciex, USA) coupled with an ExionLC™ AD ultra-high-performance liquid chromatography system (Applied Biosystems, CA, USA). The Analyst 1.7.2 software (Applied Biosystems, CA, USA) was used for data processing.
A Shim-pack GIST-HP C18-AQ column (3 µm, 2.1 × 100 mm, SHIMADZU, Japan) with a guard column (Shim-pack GIST-HP (G) C18, 3 µm, 2.1 × 10 with Cartridge (2pcs) and Holder) was used for separation. An isocratic method (100% A phase) was developed for separation. The mobile phase was water containing 0.1% formic acid and 5% isopropanol. The flow rate of the mobile phase was 0.45 mL/min. The column temperature was 35 • C, and the autosampler temperature was set at 8 • C. Acetonitrile was the wash solution of the LC system. The injection volume was 1 µL. A diverter valve was used from 1.9 min to 2.8 min.
The mass spectrometer was operated in the negative electrospray ionization (ESI) mode with multiple reaction monitoring (MRM). The transitions at m/z 116.9→72.8 for MMA and m/z 120.9→75.8 for MMA-13 C 4 were monitored for quantification. A source temperature of 450 • C and an ion spray voltage of −4500 V were used. Nitrogen gas was used as the curtain gas (CUR), nebulizer gas (GS1), auxiliary gas (GS2), and collision gas (CAD), and the pressures of the gases were set at 35, 35, 45, and low mode, respectively. The declustering potential (DP), entrance potential (EP), and collision exit potential (CXP) were set at −26 V, −12.6 V, and −11.6 V, respectively. The collision energy (CE) was set at −10.9 eV.

Method Validation
The method performance was validated according to the Clinical and Laboratory Standards Institute (CLSI) C62-A guideline [39].

Limits of Detection (LOD) and Limits of Quantification (LOQ)
Fifty microliters of standard solution of MMA (218.69 ng/g) was diluted with water to generate a series of concentrations. The diluted solutions were treated according to the sample preparation described previously. A signal-to-noise ratio (S/N) ≥ 3 with a coefficient of variation (CV) ≤ 20% for 20 injections was defined as the LOD, while an S/N ≥ 10 and a CV ≤ 20% for 20 injections was defined as the LOQ [39,40].

Analytical Precision and Recovery
Pooled serum was filtered using a disposable Corning bottle-top vacuum filter with a 0.22 µm membrane and aliquoted. Four pooled samples with different concentrations of MMA were prepared for the precision and recovery evaluation, that is, sample pools (Level 1), low-(Level 2), medium-(Level 3), and high-levels (Level 4) of QC materials described before. All samples were aliquoted into 2.0 mL Corning vials (1 mL/vial) and stored at −40 • C before analysis. Each sample was measured five times per day for five days.

Matrix Effect
The matrix addition mixing experiment was designed according to CLSI C62A [39] to assess the sample matrix effect before and after IS correction. Two different matrices were prepared: (X) the mixed standard solution of the neat analyte and IS and (Z) the extracted matrix, i.e., the matrix of serum pools (Level 1, Level 2, and Level 3) which were processed based on the established sample preparation procedure. Then, the defined amount of the mixed standard solution (i.e., X) was added to the same amount of extracted matrix (i.e., Z) to prepare the solution (Y), and the same amount of IS solution was added to the extracted matrix (i.e., Z) to prepare the solution (W). The IS in acetonitrile was previously dried under nitrogen gas and reconstituted with the primary mobile phase. The absolute matrix effects and IS-corrected matrix effects were calculated by the following equations: The absolute matrix effect (%) where A is the peak area of MMA; The IS − corrected matrix effect (%) where R is the peak area ratio of MMA to MMA-13 C 4 .

Method Applications
Fresh individual serum samples from healthy controls (n = 18, physical examination subjects with normal biochemical indicators, VB12-sufficient subjects (n = 24, VB12 > 240 pg/mL, measured by a routine immunoassay), VB12-deficient subjects (n = 13, VB12 < 240 pg/mL), patients diagnosed with anemia (n = 25), patients diagnosed with vitamin deficiency (n = 13), and patients diagnosed with colon cancer (n = 11) were randomly collected and measured in a random order. MMA levels in these populations were investigated using the established LC-MS/MS method. One positive urine sample from methylmalonic academia was measured alongside five urine samples from non-IMDs subjects.

Statistical Analysis
Statistical analysis was completed using Microsoft Excel 2016 (Microsoft Corporation, Redmond, WA, USA), SPSS 25.0 (IBM Inc., Armonk, NY, USA), and GraphPad Prism (version 8.0.0 for Windows, GraphPad Software, San Diego, CA, USA). The intra-assay, interassays, and total imprecision were calculated using one-way analysis of variance (ANOVA). Analytical performance specifications were established based on the withinsubject (CV I ) and between-subject (CV G ) biological variations of serum MMA, i.e., 7.2% and 21.1%, respectively [41]. An allowable imprecision (CV A ) = 0.75 * CV I was employed to evaluate the precision criterion (i.e., 5.5%). An allowable bias of 10.61%, derived from allowable bias = 0.375 * (CV I + CV G )1/2, was used as the bias criterion. A Mann-Whitney U analysis was used to examine the differences in MMA levels of two subjects.

Optimization of the Mobile Phase Strategy
MMA is a small polar compound with high retention properties, making the acquisition of good chromatographic behavior of MMA difficult. Searching for a suitable mobile phase component and an isocratic method cost most of the time in the method development. The first mobile phase strategy was 0.1% formic acid water solution (A phase) and methanol (B phase). An isocratic method of 50% A phase was initially chosen and the optimization determined 95% of A phase. This strategy (95% A phase) was friendly to both SA and MMA, but SA had a higher signal than MMA (see Supplementary Figure S1). Due to the strong retention properties of MMA, isopropanol was added to the A phase and its percentage was optimized from 50% to 2% (5% was adopted). The addition of isopropanol increased the MMA signal. Then, the percentage of A phase was optimized from 50% to 100% (100% was adopted).
The drying step before the LC-MS/MS analysis was optimized. The dried residue was initially reconstituted by a 5% methanol water solution containing 0.1% formic acid. Interestingly, MMA is a pH-sensitive compound because we found that when the residue was reconstituted by 5% methanol water solution containing 2% formic acid, there was no peak of MMA. Since MMA was monitored in a negative model, a low pH environment could enhance the ion suppression effects leading to a low signal. Reconstitution solutions were optimized from 5% methanol to 100% methanol without formic acid. Better chromatographic behavior, with a smaller peak width (<0.2 min), symmetrical peak, and higher signal of MMA, was obtained when the dried residue was reconstituted with 100% methanol. Thus, the drying step is no longer necessary.
Notably, since MMA is a high retention compound, the rinse solutions for the LC system were carefully optimized. Acetonitrile, methanol, and isopropanol were chosen for initial optimization. When methanol was employed for the rinse solution, obvious carryover existed after an injection of serum samples. When isopropanol was employed for rinsing, the carryover disappeared but the MMA peak in the next injection also disappeared. When acetonitrile was employed for rinsing, robust and good chromatographic behavior of MMA was obtained and no obvious carryover existed.
The injection volume was determined to be 1 µL, as the obvious peak tailing problems started to be presented when the injection volume was set to more than 1 µL.

Optimization of Sample Preparation
Protein precipitation was the main preparation step. The type of precipitation reagent and volume of the reaction system were optimized. Acetonitrile, methanol, and isopropanol were chosen for initial optimization while acetonitrile and isopropanol were excluded because their relative spiked recovery rates were less than 80%. Five reaction systems (RS) were initially explored for a good recovery rate and high signals (the large precipitation reagent volume can reduce the signals because of the dilution effects). A relatively good recovery was obtained in RS-5 (see Figure 1). because their relative spiked recovery rates were less than 80%. Five reaction systems (RS) were initially explored for a good recovery rate and high signals (the large precipitation reagent volume can reduce the signals because of the dilution effects). A relatively good recovery was obtained in RS-5 (see Figure 1).

Chromatographic Separation
The total run time of the LC-MS/MS analysis was 4.0 min per sample. The intrinsic isomer, i.e., succinic acid (SA), can be completely separated from MMA by chromatography. SA had a relatively higher signal than MMA. MMA can be significantly distinguished from SA. The retention time of MMA and SA were 2.21 min and 1.80 min, respectively. Figure 2 presents representative chromatographs of MMA and MMA-13 C4 in standard solutions (A), serum from a healthy control serum (B), and serum from a patient with VB12 deficiency (C).

Linearity, LOD, and LOQ
The average slope, intercept, and correlation relationship (R) with their 95% confidence interval (CI) obtained from 12 inconsecutive calibration curves used for analysis during two months were 0.992 (0.926 to 1.059), −0.004 (−0.012 to 0.004), and 0.997 (0.995 to 0.999), respectively. The LOD was estimated as <7.05 ng/g (0.058 µmol/L), and the CV for 20 consecutive injections (S/N > 3) was 5.24%. The LOQ was estimated as <10.41 ng/g (0.085 µmol/L), and the CV for 20 consecutive injections (S/N > 10) was 4.14%. Table 2 summarizes the precision and spiked recovery of the LC-MS/MS method at four levels of MMA on five consecutive days. The interrun CV at Level 1 (native pool), Level 2, Level 3, and Level 4 ranged from 3.09% to 5.27%; intrarun CV ranged from 1.42% to 2.69%; and the total CV ranged from 3.22% to 5.47%. All imprecision performace met the allowable precision criterion. The recoveries of MMA at the three levels were 101.51%, 92.40%, and 105.95%, respectively.

Discussion
In this study, we established an LC-MS/MS method for MMA quantification that had clear advantages over the previously established LC-MS/MS methods. It is trace, simple,

Discussion
In this study, we established an LC-MS/MS method for MMA quantification that had clear advantages over the previously established LC-MS/MS methods. It is trace, simple, fast, cheap, sensitive, accurate, robust, economic, environmentally friendly, and cost-effective for MMA measurement and is more suitable for laboratories that have large request volumes of MMA testing. These improvements were achieved using protein precipitation combined with a simple mobile phase strategy. Good chromatographic separation of MMA and its isomer was achieved by using an isocratic elution strategy, i.e., 100% A phase (deionzed water containing 0.1% formic acid and 8% isopropanol). Materials and reagents, which are complex or not always accessible, and procedures in previous methods, such as derivatization, multistep SPE, incubation, evaporation, drying, or reconstitution, were not required in this MMA quantification method. We believe this LC-MS/MS method for serum MMA is suitable and convenient for the evaluation of MMA status for VB12 deficiency and can be a reference for laboratories intending to improve their established methods as well as laboratories that plan to introduce MMA testing programs.
MMA can be a specific diagnostic biomarker of VB12 deficiency and some inborn errors of metabolism (IEM) [1,2]. It has been reported that moderately elevated MMA (over 0.4 µmol/L in serum,~46 ng/g) was an early indicator of acquired vitamin B12 insufficiency, and a massive elevation of MMA (over 40 µmol/L in serum,~4613 ng/g) could strongly indicate IMDs, e.g., methylmalonic acidemia (an IMD with a relatively high prevalence) [22]. We compared MMA levels in VB12-sufficient patients and VB12-deficient patients and observed higher MMA levels in VB12-deficient patients. Additionally, we observed that some patients diagnosed with colon cancer, anemia, diabetes, and coronary disease had elevated MMA.
One of the limitations of this study and previous studies was the lack of measurements of reference materials which have ceritificed values and uncertainty for truness. Current MMA measurements lack qualified certified reference measurement procedures and reference materials. The standard materials for calibrators and IS that we employed are certified reference materials with certified purities and uncertainties, thus, obtaining metrological traceability to the SI (System International) unit. [34,35]

Conclusions
In this study, we established a trace, simple, fast, cheap, sensitive, accurate, robust, economic, environmentally friendly, and cost-effective LC-MS/MS method for serum MMA quantification. This LC-MS/MS method can act as an easy assay and provide fast report results to evaluate the MMA status for vitamin B12 deficiency patients, and is especially applicable for large-scale MMA testing.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/diagnostics12102273/s1. Figure S1: Chromatographic separation by isocratic elution with 95% A phase (0.1% formic acid water solution) and 5% B phase (methanol). This procedure needs drying and residue reconstitution by 5% methanol. Institutional Review Board Statement: The study was conducted in accordance with the Declaration of Helsinki, and approved by the Ethics Committee of Beijing Hospital (protocol code: 2018YFC1002204, date of approval: 18 July 2022).

Informed Consent Statement:
This study was approved for the exemption from informed consent.