A magnetic resonance spectroscopy method for detecting 2-hydroxyglutarate in vivo

Abstract

The application belongs to the field of medical detection, and specifically provides a magnetic resonance spectroscopy method for detecting 2-hydroxyglutarate in vivo, comprising the following steps: (i) based on the chemical shift and J coupling characteristics of the 2HG molecular spin system, the selective detection of the 2HG signal near 2.22 ppm is realized through an echo time optimized magnetic resonance spectroscopy sequence; (ii) a basic set containing 2HG and common brain metabolites is established based on the evolution of the metabolite spin system; (iii) the collected spectroscopy data is preprocessed, and metabolite fitting and quantification are realized based on the basic set; (iv) the post-processing data is quality controlled and 2HG false positive analysis is performed to generate a structured report. The method has the advantages of high sensitivity, good stability and strong expansibility, and can realize the non-invasive detection of 2HG.

Description

Technical Field

[0001] This invention belongs to the field of medical testing, specifically providing a magnetic resonance spectroscopy method for detecting 2-hydroxyglutaric acid in vivo. Background Technology

[0002] 2-hydroxyglutarate (2HG) is a specific metabolite that accumulates abnormally due to mutations in the isocitrate dehydrogenase (IDH) gene. Its levels can be used to identify IDH mutation status, differentiate tumor molecular subtypes, assess disease prognosis, and monitor treatment response, playing a significant role in the precision diagnosis and treatment of brain tumors and hematological malignancies. Current 2HG detection methods primarily rely on tissue biopsy combined with mass spectrometry or liquid chromatography for quantitative analysis. While offering high measurement accuracy, these methods are invasive, posing risks such as bleeding and infection, making them unsuitable for follow-up dynamic monitoring. Furthermore, these methods are dependent on sampling location and susceptible to the spatial heterogeneity of tumors, failing to comprehensively reflect the overall metabolic status of the tumor.

[0003] Magnetic resonance spectroscopy (MRS) is a non-invasive technique for acquiring the concentration and spatial distribution of metabolites in vivo, directly characterizing tissue metabolic status and avoiding exogenous damage associated with biopsies. However, in conventional 1H MRS, 2HG signals exhibit significant spectral overlap with metabolites such as glutamate and glutamine, making specific separation and reliable quantification of 2HG difficult. Furthermore, current spectral analyses often lack a basic set of metabolites matched to specific acquisition protocols, leading to large fitting errors for 2HG and other metabolites. Simultaneously, the lack of systematic quality control and false-positive analysis procedures makes it difficult to guarantee the stability and reliability of the test results.

[0004] Given the aforementioned limitations, there is an urgent need for a non-invasive, targeted, and highly sensitive 2HG magnetic resonance spectroscopy detection method. This method should improve the detection specificity and quantitative accuracy of 2HG, construct a basic set matched to specific field strengths, metabolite categories, sequence types, and sequence parameters, and possess a robust quality control and false-positive analysis mechanism to ensure data reliability. This would provide stable and efficient technical support for the non-invasive detection, clinical diagnosis, and scientific research of 2HG. Summary of the Invention

[0005] To address the aforementioned issues, this invention proposes an in vivo 2HG magnetic resonance spectroscopy detection method for non-diagnostic purposes. The method standardizes and optimizes the 2HG measurement process from aspects such as data acquisition, basic set generation, spectral post-processing, and quality control analysis, thereby improving the accuracy and stability of the measurement results.

[0006] This invention provides a magnetic resonance spectroscopy method for detecting 2-hydroxyglutaric acid in vivo, comprising the following steps: v. Based on the chemical shift and J coupling characteristics of the 2HG molecular spin system, selective detection of 2HG signals near 2.22 ppm is achieved through a magnetic resonance spectroscopy sequence optimized by echo time; vi. Establish a basic set including 2HG and brain metabolites based on the evolution of metabolite spin systems; vii. Preprocess the acquired spectral data and perform metabolite fitting and quantification based on the basic set; viii. Perform quality control and 2HG false positive analysis on the post-processed data, and generate a structured report.

[0007] In one embodiment, the method provided by the present invention, in step (i), the 2HG molecular spin system includes one methine proton and four protons on two methylene groups; the echo time is an optimized magnetic resonance spectral sequence determined by simulation calculation and experiment of the 2HG spin system, used to present the 2HG characteristic signal at 2.22 ppm and suppress the signal of brain metabolites, wherein the brain metabolites are glutamate and glutamine.

[0008] In one embodiment, the method provided by the present invention includes a volumetric external suppression module, a water suppression module, and a signal acquisition module for the echo time optimized magnetic resonance spectral sequence. Further, the volumetric external suppression module suppresses extravoxel metabolite and water molecule signals by applying multiple saturation pulses. The water suppression module employs CHESS, WET, or VAPOR water suppression techniques. The signal acquisition module consists of multiple layer-selective radio frequency pulses and gradient pulses.

[0009] In one embodiment, the method provided by the present invention uses a magnetic resonance spectrum sequence optimized for echo time, which is either a spin echo sequence or a stimulated echo sequence.

[0010] In one embodiment, the method provided by the present invention uses PRESS, sLASER, or LASER as the spin echo sequence and STEAM as the stimulated echo sequence.

[0011] In one embodiment, the method provided by the present invention sets the echo time in the range of 80 ms to 110 ms; further, for the spin echo sequence, 97 ms is preferred.

[0012] In one embodiment, the method provided by the present invention, in step (ii), the basic set of 2HG and brain metabolites is determined by the magnetic resonance field strength, spectral sequence type, experimental parameters, and metabolite type; the spin system of each metabolite undergoes spin evolution based on its proton chemical shift and J-coupling characteristics, combined with the sequence type and experimental parameters, to form a metabolite template for fitting actual magnetic resonance spectral data; further, the magnetic resonance field strength is 3.0; the spectral sequence type is PRESS, sLASER, LASER, or STEAM; the experimental parameters include echo time, repetition time, pulse shape of excitation and refocusing pulses, frequency bandwidth, flip angle, gradient intensity, and action time.

[0013] In one embodiment, the method provided by the present invention includes spectral data preprocessing in step (iii), which includes frequency correction, phase correction, baseline fitting, and residual water peak removal; the preprocessed spectral data is fitted based on the base set to obtain the concentrations of 2HG and other metabolites and their fitting error ranges.

[0014] In one embodiment, the method provided by the present invention includes quality control of post-processing data in step (iv), which includes the evaluation of water peak half-width, signal-to-noise ratio, and metabolite fitting error; after the quality control is passed, false positive analysis is performed, including analysis of the selected voxel position, lactate peak, and glutamate intensity; after the false positive analysis is passed, a structured report is generated, which includes basic patient information, metabolite concentration, quality control results, and diagnostic recommendations.

[0015] On the other hand, the method provided by the present invention is applied to the non-invasive detection of 2HG in brain tumor tissue, and is used to identify the IDH gene mutation status, assist in clinical diagnosis, and evaluate the treatment effect.

[0016] S1. This invention provides a magnetic resonance spectroscopy method for detecting 2-hydroxyglutaric acid (2HG) in vivo, specifically including the following steps: (1) Based on the chemical shift and J coupling characteristics of the 2HG molecular spin system, selective detection of 2HG signal near 2.22 ppm is achieved through a magnetic resonance spectroscopy sequence optimized by echo time; (2) A basic set including 2HG and common brain metabolites is established based on the evolution of the metabolite spin system; (3) The spectral data collected first is preprocessed, and then the metabolites are fitted and quantified based on the basic set; (4) The post-processed data is quality controlled and 2HG false positive analysis is performed to generate a structured report.

[0017] S2. The 2HG molecular spin system includes one methine proton and four protons on two methylene groups, with characteristic spectral peaks formed at approximately 4.02 ppm, 2.27 ppm, 2.22 ppm, 1.98 ppm, and 1.83 ppm for proton chemical shifts. The optimized magnetic resonance spectral sequence can be a spin echo sequence, such as PRESS, sLASER, or LASER, or a stimulated echo sequence, such as STEAM. The spectral sequence is obtained by simulating the 2HG five-spin system and selecting an echo time that brings the 2HG signal intensity close to its maximum at approximately 2.23 ppm while significantly reducing the signals of metabolites such as glutamate, glutamine, and glutathione. The echo time is set in the range of 80 ms to 110 ms, and preferably 97 ms for spin echo sequences.

[0018] S3. The echo time-optimized magnetic resonance spectral sequence includes a volumetric external suppression module, a water suppression module, and a signal acquisition module. The volumetric external suppression module suppresses extravoxel metabolite and water molecule signals by applying multiple saturation pulses. The water suppression module uses water suppression techniques such as CHESS, WET, or VAPOR to suppress water signals within the region of interest. The signal acquisition module consists of multiple layer-selective radio frequency pulses and gradient pulses, used to acquire magnetic resonance spectral signals of metabolites within the region of interest.

[0019] S4. The basic set of 2HG and common brain metabolites is determined by the magnetic resonance field strength, spectral sequence type, experimental parameters, and metabolite type. Specifically, under given magnetic resonance main magnetic field strength (e.g., 3.0 T or 7.0 T), selected spectral sequence type (e.g., PRESS, sLASER, LASER, or STEAM), and corresponding experimental parameters (including echo time TE, repetition time TR, pulse shape, bandwidth, flip angle, gradient intensity, and duration of excitation and refocusing pulses), the spin systems of 2HG and various common brain metabolites are modeled and simulated. The spin system of each metabolite may include a single or multiple spin subsystems, defined according to its proton chemical shift, number of protons, and J coupling constant between protons. Based on this, using density matrix or equivalent quantum mechanical spin evolution methods, under the selected spectral sequence pulse timing and experimental parameters, the time-domain evolution simulation of the spin system of each metabolite is performed to obtain the corresponding ideal free induction decay signal (FID) or its spectrum. Subsequently, linewidth broadening, phase adjustment, and noise level settings can be applied to the simulated signal according to actual experimental conditions to obtain standardized metabolite spectral templates that match the actual acquisition conditions. Using the above method, a set of basic spectral libraries corresponding to specific field strengths and sequence parameters is generated for 2HG and common brain metabolites such as N-acetylaspartate, creatine, phosphocreatine, choline, inositol, glutamate, glutamine, γ-aminobutyric acid, lactate, and glycine; this is the basic set. The basic set can be stored as standardized metabolite templates in the frequency or time domain and a unified normalization strategy (normalized by unit concentration) is used to fit the actual magnetic resonance spectral data. In subsequent data analysis, the preprocessed measured spectra are linearly or nonlinearly combined with this basic set to obtain the concentration estimates of 2HG and other metabolites and the corresponding fitting errors, thereby achieving accurate quantitative analysis based on spin system evolution.

[0020] S5. The quality control of the post-processed data includes the evaluation of indicators such as the half-width at half-maximum (WHM) of the water peak, the signal-to-noise ratio (SNR), and the fitting error of metabolites. First, the WHM of the metabolite with the highest concentration in the spectrum (usually choline or creatine) is measured to evaluate the local magnetic field homogeneity and spectral linewidth. If the WHM is greater than a preset threshold, the data is not included in subsequent analysis. Second, the overall SNR of the spectrum within the target voxel is calculated. If the SNR is lower than a preset lower limit, the data quality is considered insufficient to support reliable quantification. After the basic set is fitted, the fitting residuals and fitting errors of 2HG and other major metabolites are statistically analyzed. If the fitting error exceeds the allowable range, the quality control fails. After the quality control passes, false positive analysis is performed. False positive analysis includes checking the rationality of the selected voxel location to ensure that the voxel is mainly located in the tumor parenchyma rather than in necrotic areas, cystic areas, or areas of obvious hemorrhage or calcification; analyzing the spectral morphology and relative intensity of metabolites such as glutamate and glutamine to exclude cases where spectral line overlap is mistakenly identified as 2HG signals. In addition, observe the lactate signal intensity. If the lactate signal is too high, consider that the selected voxel may contain a cavity, and there may be other metabolites (such as succinic acid) signals in the cavity, which may be mistaken for 2HG signals. If the voxel localization, lactate peak characteristics, and glutamate signal intensity are consistent with the 2HG positive manifestation, and there is no obvious abnormality in spectral fitting, then the false positive analysis is passed.

[0021] S6. If both quality control and false positive analyses pass, the system generates a structured report. This report includes at least the patient's basic information (such as patient ID, gender, age, etc.), scan and sequence parameters, concentrations or relative concentrations of each metabolite (including 2HG) and their error estimates, quality control results and judgment conclusions, and provides quantitative or semi-quantitative diagnostic suggestions based on 2HG levels and other metabolite profile characteristics for clinical image reading and follow-up comparison.

[0022] Compared with the prior art, the present invention has the following beneficial effects: (1) This invention provides an in vivo magnetic resonance spectroscopy detection method for non-diagnostic purposes. Compared with traditional invasive detection methods that rely on tissue biopsy and mass spectrometry analysis, this method is non-invasive, can be used to assess the metabolic status of tumors, avoids biopsy-related risks, and is suitable for long-term follow-up and dynamic monitoring.

[0023] (2) By optimizing the echo time, the present invention can more accurately separate and quantify the 2HG signal, avoiding the measurement error caused by spectral overlap and signal interference in conventional MRI spectra, thereby providing more reliable 2HG concentration measurement results.

[0024] (3) The present invention includes data acquisition, basic set creation, data post-processing, data quality control and false positive analysis modules. Compared with conventional MRS technology, it has significant advantages in the specific identification of 2HG signals, the accuracy of metabolite quantification, and the control of result quality and false positives. It can realize high sensitivity, repeatable and non-invasive detection of 2HG in vivo, and provide a more reliable technical means for the metabolic assessment of IDH mutation-related tumors. Attached Figure Description

[0025] Figure 1 Ha, Hb, Hb', Hc, and Hc' represent the five protons on the 2HG methine and the two sets of methylene groups, respectively, forming a five-spin system. The chemical shifts of each proton are 4.02 ppm, 1.98 ppm, 1.83 ppm, 2.22 ppm, and 2.27 ppm, respectively.

[0026] Figure 2 A schematic diagram of the pulse sequence with optimized echo time is shown, where Gx, Gy, and Gz represent the pulse gradient channels in the x, y, and z directions, respectively, and 90x, 180y, and 180y are the flip angle and phase of the pulse used for layer selection, respectively. The sequence has an echo time of 97 ms.

[0027] Figure 3 The images show the ¹H MRS spectra of 2HG and Glu obtained from PRESS sequences at different echo times in Example 1. (a) Spectra of 2HG, Glu, and a mixture of the two (2HG+Glu) obtained from PRESS sequences with an echo time of 30 ms; (b) Simulated spectra of 2HG, Glu, and 2HG+Glu obtained from PRESS sequences with an echo time of 97 ms.

[0028] Figure 4 A schematic diagram of the metabolite base set used in Embodiment 1 of the present invention.

[0029] Figure 5 This is a schematic diagram of the 2HG magnetic resonance spectral structure report involved in this invention.

[0030] Figure 6 This is a flowchart illustrating the specific implementation of the method of the present invention; Figure 7 The image shows the spectra acquired in a 65-year-old patient with an IDH-mutant brain tumor using a PRESS sequence with time-optimized echo in Embodiment 2 of the present invention. Figure 8 This is a spectral image acquired in a 60-year-old wild-type brain tumor patient using a PRESS sequence with echo time optimization in Embodiment 3 of the present invention.

[0031] Figure 9 The ¹H MRS results of a 68-year-old patient with an IDH-mutant brain tumor were acquired using a TE = 97 ms PRESS sequence before (a), 75 days after (b), and 154 days after (c). Detailed Implementation

[0032] The present invention will be further described below with reference to specific embodiments. It should be noted that the following embodiments are provided to enable those skilled in the art to better understand the present invention, and are not intended to limit the invention. Unless otherwise specified, the reagents and instruments used below are all commercially available.

[0033] Example 1 illustrates the design process of PRESS sequences with echo time optimization and the method for constructing a basic set of 2HG and common brain metabolites. Figures 1-6 2HG consists of a five-spin system comprising one methine proton and four protons on two sets of methylene groups, such as... Figure 1 As shown, the chemical shifts of each proton are approximately 4.02 ppm, 1.98 ppm, 1.83 ppm, 2.22 ppm, and 2.27 ppm, respectively. The J-coupling parameters between protons are described in relevant literature. In this embodiment, a PRESS sequence (e.g., under a 3.0 T magnetic field) is used. Figure 2 As shown in the figure, the parameters are as follows: TR = 2000 ms, TE = 97 ms; the excitation pulse and the two refocusing pulses are both sinc pulses with pulse widths of 1 ms and 2 ms, respectively; the time intervals from the excitation pulse to the two refocusing pulses are 5 ms and 43.5 ms, respectively; the linewidth is set to 4 Hz; the sampling bandwidth is 1200 Hz; and the number of sampling points is 1024.

[0034] Since glutamate (Glu) is the metabolite most likely to overlap with the 2HG signal at approximately 2.22 ppm, the spectra of Glu, 2HG, and their mixtures were simulated at different echo times, such as... Figure 3 As shown. Figure 3 a shows the simulation results of a PRESS sequence with an echo time of 30 ms. In the spectrum where 2HG and Glu coexist, the signals of 2HG and Glu overlap significantly near 2.22 ppm, making them difficult to distinguish. Figure 3 Figure b shows the simulation results of the PRESS sequence with an echo time of 97 ms. The 2HG signal shows a positive peak near 2.22 ppm, while the Glu signal shows an inverted peak, which significantly improves the separation between 2HG and Glu, facilitating subsequent fitting and quantitative analysis.

[0035] To fit and quantify the signals of 2HG and other brain metabolites, spin system models of 2HG and common brain metabolites were established (the spin system of a metabolite can be a single spin system or include multiple spin subsystems). Time-domain evolution simulations were performed using PRESS pulse sequences to obtain the standardized spectral signals of each metabolite. All simulated signals were normalized to unit concentration, forming a model as follows: Figure 4 The basic set of metabolites shown is included. This basic set contains metabolites such as 2HG, NAA, Ins, Gly, Gln, GSH, GABA, PCr, Lac, Glu, Glc, GPC, and Cr.

[0036] Based on metabolite fitting and quantification, information such as the concentrations of 2HG and other metabolites, signal-to-noise ratio, spectral half-width, and fitting uncertainty are output in a structured report format for clinical image reading and follow-up comparison (e.g., Figure 5 (As shown). In summary, this embodiment uses a PRESS sequence with an echo time of 97 ms as an example to demonstrate the overall process of the 2HG magnetic resonance spectroscopy detection method (as shown). Figure 6 As shown in the figure, the process includes steps such as data acquisition, spectral preprocessing, basic set fitting, quality control, and 2HG false positive analysis, realizing a standardized processing link from data acquisition to report generation.

[0037] Example 2 illustrates the specific application of the time-optimized PRESS sequence of the present invention in patients with IDH-mutant brain tumors, and the process of automatically generating 2HG-related structured reports based on the fitting results. Figure 7 .

[0038] A 65-year-old brain tumor patient with IDH mutation confirmed by surgical pathology and molecular testing was selected. Routine MRI examination was performed first, including T1WI, T2WI, and FLAIR sequences. The extent of the tumor lesion was determined on the axial FLAIR image, and the monomeric chromosomal sample was placed within the tumor parenchyma, avoiding areas of significant necrosis, cystic degeneration, and hemorrhage as much as possible.

[0039] On a Siemens 3.0 T magnetic resonance system, monomorphic ¹H MRS acquisition was performed using the PRESS sequence, with a TR of 2000 ms, a TE of 97 ms, a sampling bandwidth of 1200 Hz, a voxel size of 2 × 2 × 2 cm³, 1024 sampling points, and a signal averaging frequency of 128. The acquired raw spectra underwent preprocessing including frequency correction, phase correction, residual water peak removal, baseline fitting, and linewidth unification to obtain the following results: Figure 7 The measured spectrum is shown by the black curve in the lower left corner; fitting is performed using the basic set constructed in Example 1 to obtain the following... Figure 7The red curve in the middle shows the fitting results and the spectra of each metabolite, with the fitting spectra of Gly and 2HG shown separately.

[0040] After the fitting is completed, the system performs quality control evaluation on the data, such as... Figure 7 As shown, the "Data Quality Control" module has an SNR of 26, FWHM of 3.7 Hz, and a %SD of tCr of 2%, all meeting the preset threshold requirements, thus the quality control is considered passed. The "Metabolite Concentration" module shows: tCho concentration is 4.04 mM ( / tCr ratio 0.57), tNAA concentration is 1.43 mM, tCr concentration is 7.15 mM ( / tCr = 1), Glu concentration is 1.77 mM, Gly concentration is 1.88 mM, Lac concentration is 2.93 mM, and 2HG concentration is 5.75 mM. The fitted %SD of the main metabolites are all within the range of 1%–14%. Subsequent 2HG false-positive analysis and comprehensive diagnostic assessment showed that the 2HG concentration was significantly higher than 0.5 mM, and the fitting error (%SD) was less than 50%, meeting the quantitative criteria for 2HG positivity. Simultaneously, lactate and glutamate concentrations were at relatively low levels, with no abnormal increases, ruling out false positives caused by high lactate peaks or overlapping Glu / Gln spectral lines. The false-positive analysis result was "passed". Further calculation of metabolite ratios showed tNAA / tCr = 0.2 and Gly / 2HG = 0.43, indicating a relative decrease in NAA, an increase in choline, and a synchronous increase in 2HG and Gly, consistent with the metabolic characteristics of IDH-mutant tumors. Based on the above quantitative results, quality control indicators, and ratio analysis, the system presented a "2HG positive" diagnostic conclusion in the structured report.

[0041] Example 3: A 60-year-old brain tumor patient with IDH wild-type tumor, confirmed by postoperative pathology and gene testing, was selected. The extent of the tumor lesion was determined on axial FLAIR sequences. A voxel was placed within the tumor parenchyma, avoiding areas of significant necrosis, cystic degeneration, and hemorrhage as much as possible. The voxel placement is as follows: Figure 8 As shown in the top left corner.

[0042] On a Siemens 3.0 T magnetic resonance system, monomorphic ¹H MRS acquisition was performed using the PRESS sequence, with a TR of 2000 ms, a TE of 97 ms, a sampling bandwidth of 1200 Hz, a voxel size of 2 × 2 × 2 cm³, 1024 sampling points, and a signal averaging frequency of 128. The acquired raw spectra underwent preprocessing including frequency correction, phase correction, residual water peak removal, baseline fitting, and linewidth unification to obtain the following results: Figure 8 The measured spectrum is shown by the black curve in the lower left corner; fitting is performed using the basic set constructed in Example 1 to obtain the following... Figure 8The red curve in the middle shows the fitting results and the spectra of each metabolite, with the fitting spectra of Gly and 2HG shown separately.

[0043] After the fitting is completed, the system performs quality control evaluation on the data, such as... Figure 8 As shown, the "Data Quality Control" module has an SNR of 14, FWHM of 4.9 Hz, and tCr %SD of 3%, all meeting the preset threshold requirements, indicating that quality control has passed. The "Metabolite Concentration" module shows: tCho concentration is 3.18 mM ( / tCr ratio 0.48), tNAA concentration is 4.78 mM ( / tCr = 0.72), tCr concentration is 6.66 mM ( / tCr = 1), Glu concentration is 4.86 mM, Gly concentration is 0.59 mM, and the fitted concentrations of Lac and 2HG are both 0 mM. Subsequent comprehensive diagnostic evaluation revealed that the 2HG concentration <0.5 mM and the fitting error >50%, failing to meet the 2HG positive criteria; tNAA / tCr = 0.72 indicates a decrease in NAA, consistent with the metabolic characteristics of wild-type tumors. Based on the combined quantitative results, ratio analysis, and quality control status, the structured report concludes that the diagnosis is 2HG negative.

[0044] Example 4 presents the 2HG measurement results of a 68-year-old brain tumor patient with an IDH mutation confirmed by postoperative pathology and molecular testing, measured on days before drug treatment, day 75, and day 154 after treatment. All three examinations were performed on a Siemens 3.0 T MRI system using the PRESS sequence of this invention with a TE = 97 ms, TR = 2000 ms, a sampling bandwidth of 1200 Hz, 1024 sampling points, and a signal averaging of 128 times. Voxels were placed within the tumor parenchyma to maintain consistency in voxel position and volume across the three follow-up visits, ensuring comparability for longitudinal comparisons.

[0045] After preprocessing, the raw spectral data were fitted using the basic set constructed in Implementation 1 to obtain the concentration of each metabolite and the fitting uncertainty. At the same time, the water peak half width at half maximum, signal-to-noise ratio (SNR), and tCr fitting %SD were evaluated for quality control. All three scans met the preset quality control thresholds, and the quality control was passed.

[0046] The fitting results of Gly and 2HG for three scans are as follows: Figure 9 As shown. Baseline ( Figure 9 At time a), the Gly concentration was 0.38 mM and the 2HG concentration was 3.36 mM, indicating a significant abnormal accumulation of 2HG within the lesion, consistent with an IDH mutation-related metabolic phenotype. On day 75 of treatment ( Figure 9(b) Under conditions where the voxel and acquisition parameters were basically consistent, a follow-up examination showed that the Gly concentration decreased to 0 mM, while the 2HG concentration decreased to 2.63 mM, a decrease of approximately 22% from the baseline, indicating an early decreasing trend in intratumoral 2HG levels. Day 154 of treatment ( Figure 9 c) At the second follow-up, the Gly concentration was 0.45 mM, and the 2HG concentration further decreased to 1.50, a decrease of more than 50% from the baseline, showing a sustained and significant downregulation of 2HG.

[0047] Compared with the technique in "A Preliminary Study on the Efficacy of 3.0 T Monomeric Magnetic Resonance Spectroscopy Imaging for In Vivo and In Vivo Detection of 2-Hydroxyglutarate", the pulses used are similar and the optimized TE time is similar; however, the article has not yet mentioned the method for judging false positives of 2HG; in addition, the base set does not include tumor metabolite components, such as glycine and alanine molecules, making it difficult to guarantee the accuracy of the fitted metabolite concentrations.

[0048] Compared to the technique described in "Detection of 2-Hydroxyglutarate in IDH-Mutated Glioma Patients by In Vivo Spectral-Editing and 2DCorrelation Magnetic Resonance Spectroscopy," the TE-optimized PRESS technique mentioned in this patent achieves over 40% higher detection efficiency for 2HG in MEGA-sLASER sequences; furthermore, it offers a shorter detection time than 2D COSY technology. Based on the quantitative results from three follow-up visits and stable quality control indicators, it is evident that, under controllable systematic errors, 2HG concentration exhibits a stepwise decreasing trend with treatment progress, reflecting the gradual suppression of IDH mutation-related metabolic abnormalities during treatment. This embodiment demonstrates that the 2HG quantification method of this invention, based on echo-time optimized PRESS sequences and a matched baseline set, can provide sensitive and reproducible metabolic endpoints in longitudinal follow-ups and can serve as an objective biomarker for evaluating the efficacy of drugs for IDH-mutant brain tumors.

Claims

1. A magnetic resonance spectroscopy method for detecting 2-hydroxyglutaric acid in vivo, characterized in that, Includes the following steps: i. Based on the chemical shift and J coupling characteristics of the 2HG molecular spin system, selective detection of 2HG signals near 2.22 ppm is achieved through a magnetic resonance spectroscopy sequence optimized by echo time; ii. Establish a basic set including 2HG and brain metabolites based on the evolution of metabolite spin systems; iii. Preprocess the collected spectral data and perform metabolite fitting and quantification based on the basic set; iv. Perform quality control and 2HG false positive analysis on the post-processed data, and generate a structured report.

2. The method according to claim 1, characterized in that, In step (i), the 2HG molecular spin system includes one methine proton and four protons on two methylene groups; the echo time is an optimized magnetic resonance spectral sequence determined by simulation calculation and experiment of the 2HG spin system, used to present the 2HG characteristic signal at 2.22 ppm and suppress the signal of brain metabolites, wherein the brain metabolites are glutamate and glutamine.

3. The method according to claim 2, characterized in that, The echo time-optimized magnetic resonance spectral sequence includes a volumetric external suppression module, a water suppression module, and a signal acquisition module. The volumetric external suppression module suppresses extravoxel metabolites and water molecule signals by applying multiple saturation pulses. The water suppression module employs CHESS, WET, or VAPOR water suppression techniques. The signal acquisition module consists of multiple layer-selective radio frequency pulses and gradient pulses.

4. The method according to claim 3, characterized in that, The sequence is a magnetic resonance spectrum sequence optimized for the echo time, which is either a spin echo sequence or a stimulated echo sequence.

5. The method according to claim 4, characterized in that, The spin echo sequence is PRESS, sLASER, or LASER; the stimulated echo sequence is STEAM.

6. The method according to claim 5, characterized in that, The echo time is set to 97 ms.

7. The method according to claim 1, characterized in that, In step (ii), the basic set of 2HG and brain metabolites is determined by the magnetic resonance field strength, spectral sequence type, experimental parameters, and metabolite type. The spin system of each metabolite undergoes spin evolution based on its proton chemical shift and J-coupling characteristics, combined with the sequence type and experimental parameters, to form a metabolite template for fitting actual magnetic resonance spectral data. The magnetic resonance field strength is 3.

0. The spectral sequence type is PRESS, sLASER, LASER, or STEAM. The experimental parameters include echo time, repetition time, pulse shape of excitation and refocusing pulses, frequency bandwidth, flip angle, gradient intensity, and duration of action.

8. The method according to claim 1, characterized in that, The spectral data preprocessing in step (iii) includes frequency correction, phase correction, baseline fitting, and residual water peak removal; the preprocessed spectral data is fitted based on the base set to obtain the concentrations of 2HG and other metabolites and their fitting error ranges.

9. The method according to claim 1, characterized in that, The quality control of post-processed data in step (iv) includes the evaluation of water peak half-width, signal-to-noise ratio, and metabolite fitting error; after the quality control is passed, false positive analysis is performed, including analysis of the selected voxel position, lactate peak, and glutamate intensity; after the false positive analysis is passed, a structured report is generated, which includes basic patient information, metabolite concentration, quality control results, and diagnostic recommendations.

10. The method according to claim 1, characterized in that, The method is applied to the non-invasive detection of 2HG in brain tumor tissue, and is used to identify the IDH gene mutation status, assist in clinical diagnosis, and evaluate treatment efficacy.

Images