Primary aldosteronism typing diagnosis model and construction method

Abstract

The invention provides a construction method of a primary aldosteronism typing diagnosis model. The construction method comprises the following steps: a, collecting clinical data: taking 24-hour urine aldosterones, the lowest blood potassium level and typical adenoma shown in imageology as indexes; and b, the primary aldosteronism can be subjected to typing diagnosis through a prediction model established by drawing a working characteristic curve of a subject. The invention further provides the primary aldosteronism typing diagnosis model constructed by the construction method of the primary aldosteronism typing model, and the primary aldosteronism typing diagnosis model is constructed by the construction method of the primary aldosteronism typing model. It is found that 24 h-UAC of a unilateral PA patient is obviously higher than 24 h-UAC of a bilateral PA patient, and the kit can be used for typing diagnosis of PA; a prediction model established by combining the three indexes of 24h-UAC, the minimum blood potassium level and typical adenoma shown in imageology has relatively high diagnosis efficiency on typing diagnosis of PA, and can provide help for clinical typing of PA.

Description

Technical Field

[0001] The invention relates to a primary aldosteronism typing diagnosis model, and belongs to the field of diagnosis. Background Art

[0002] Primary aldosteronism (PA) is a secondary hypertension caused by the autonomous secretion of aldosterone. Compared with primary hypertension, long-term and large-scale exposure of PA patients to excessive aldosterone will cause oxidative stress, inflammation, endothelial dysfunction, fibrosis, and significantly increase the risk of cardiovascular and cerebrovascular events and target organ damage such as kidneys. PA is divided into unilateral PA (UPA) and bilateral PA (BPA). Correct classification diagnosis determines the choice of treatment. Currently, the classification diagnosis method recommended by multiple consensuses is bilateral adrenal venous sampling (AVS). However, AVS is an invasive and invasive procedure with high cost and difficulty, making it difficult to popularize on a large scale. Although a few studies have found that CXC chemokine receptor 4 (CXCR4) imaging may be helpful in diagnosing unilateral aldosteroneoma, there is no diagnostic standard for patients with bilateral lesions shown on CT, and it is difficult to promote on a large scale. For example, the use of PET-CT also faces radiation risks. Therefore, it is urgent to explore simpler, more convenient and non-invasive typing methods in clinic. In the past, a few studies have used immunoassays to detect the 24-hour urinary aldosterone concentration (24h-UAC) of PA patients, and found that the 24h-UAC of UPA patients was significantly higher than that of BPA patients, and the area under the ROC curve (AUC) for distinguishing unilateral and bilateral lesions was as high as 0.898-0.921. However, during the immunoassay detection process, antibodies and other steroid hormones or metabolites are prone to cross-reaction, which can easily interfere with the test results and cause false positives. At present, liquid chromatography tandem mass spectrometry (LC-MS / MS) has been recognized at home and abroad as the most accurate method for detecting steroid hormones. It has the advantages of high sensitivity, high specificity, and high throughput, and the detection results are more accurate than immunoassays. [1] .

[0003] Reincke et al. analyzed the univariate related factors of mortality in PA patients and found that adrenalectomy can significantly reduce the all-cause mortality of PA patients compared with aldosterone receptor antagonist treatment. Therefore, correct classification diagnosis is the key to determining precise treatment and improving prognosis. However, the classification diagnosis of PA has always been a difficult point in clinical practice. Adrenal CT, AVS, and CXCR4 imaging can all be used as methods for classification diagnosis, but they all have shortcomings. In recent years, exploring simple, convenient and non-invasive classification methods has always been a hot topic in clinical research, especially using clinical specific indicators for classification diagnosis is a more ideal choice.

[0004] 24h-UAC is the total amount of aldosterone secreted in urine over 24 hours. It can avoid the influence of factors such as body position and circadian rhythm. In theory, it can better reflect the overall aldosterone secretion of the body in a day than PAC at a single time point. [2] It is a clinical specific indicator that reflects the aldosterone secretion status of PA patients, and it is also a clinical indicator for PA typing diagnosis that is worthy of exploration. At present, the main clinical aldosterone detection method is immunoassay, including radioimmunoassay and chemiluminescence immunoassay. However, the immunoassay is based on the reaction principle of specific binding of antigen and antibody, and is easily interfered by cross-reactions, resulting in detection values ​​that are higher than the actual values. The LC-MS / MS method has the characteristics of high sensitivity and high specificity, and is the "gold standard" for the determination of small molecule hormones, but its clinical application is still in its early stages. At present, most of the research on 24h-UAC focuses on the screening of PA, and fewer explore its typing diagnostic value. In addition, most of the research methods use immunoassays, and only a very small number use LC-MS / MS. Ma Wenjun, Union Hospital [3] et al. used LC-MS / MS to detect 24h-UAC and found that its optimal cut-off value for diagnosing PA was 7.13μg / 24h, but the article did not further explore its value in typing diagnosis. [4] et al. also used LC-MS / MS to detect 24h-UAC and found that 24h-UAC can provide a reference for clinical screening and diagnosis of PA. If combined with renin detection, it can provide screening and diagnostic value equivalent to ARR. In addition, it was found that 24h-UAC and urine aldosterone-renin ratio can be used as good predictive indicators for PA typing diagnosis. However, the research results of this article cannot be directly applied to the clinical diagnosis and typing of PA, and cannot directly distinguish between unilateral and bilateral PA. Summary of the invention

[0005] The present invention discloses the value of 24h-UAC detected by LC-MS / MS and its related diagnostic model for the typing diagnosis of Chinese PA patients.

[0006] The present invention provides a method for constructing a primary aldosteronism typing diagnosis model, which comprises the following steps:

[0007] a. Collect clinical data: 24-hour urine aldosterone, lowest blood potassium level, and imaging manifestations are used as typical adenoma indicators;

[0008] b. The prediction model established by drawing the receiver operating characteristic curve is used to diagnose the primary aldosteronism.

[0009] Wherein, the prediction model formula described in step b is:

[0010] F = 2.097 + 0.166 × 24h-UAC - 1.660 × lowest serum potassium level + 1.310 × imaging findings were typical adenoma;

[0011] Among them, F is the regression equation logist(P), which is the quantitative relationship between the dependent variable primary aldosteronism typing diagnosis and the three independent variables of 24-UAC, lowest blood potassium level, and imaging manifestation of typical adenoma established by regression analysis;

[0012] 24h-UAC is the total amount of urinary aldosterone in 24 hours measured by liquid chromatography tandem mass spectrometry, expressed in μg / 24h;

[0013] The unit of the lowest blood potassium level is mmol / L;

[0014] If the imaging findings were typical adenoma, yes = 1, no = 0.

[0015] Specifically, it includes the following steps:

[0016] a. 24-UAC determined by liquid chromatography tandem mass spectrometry;

[0017] b. The lowest value of blood potassium level detected;

[0018] c. Evaluate imaging findings;

[0019] d. Calculate the F value according to the prediction model formula;

[0020] e. Primary aldosteronism is classified according to the F value.

[0021] The present invention also provides a typing model constructed by the method for constructing a typing model of primary aldosteronism.

[0022] The present invention provides a method for predicting the typing of primary aldosteronism using the primary aldosteronism typing diagnosis model, wherein the F value is greater than or equal to 0.704, indicating unilateral primary aldosteronism, which can be treated surgically. The F value is less than 0.704, indicating bilateral primary aldosteronism, which should be treated with medication.

[0023] The present invention found that the 24h-UAC of patients with unilateral PA was significantly higher than that of patients with bilateral PA, which can be used for the classification diagnosis of PA; the prediction model established by combining the three indicators of 24h-UAC, the lowest blood potassium level, and the imaging manifestation of typical adenoma has a high diagnostic efficacy for the classification diagnosis of PA, which may provide assistance for the clinical classification of PA. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 : Comparison of 24h-UAC between UPA and BPA patients;

[0025] Figure 2 :ROC curve of 24h-UAC for PA typing diagnosis;

[0026] Figure 3 :Comparison of ROC curves between 24h-UAC and the prediction model combining 24h-UAC, lowest serum potassium level, and typical adenoma on imaging;

[0027] Figure 4 : ROC curves of the model of the present invention, the Chinese modified Kupers score and the CONPASS model. DETAILED DESCRIPTION

[0028] Example 1 Establishment of the primary aldosteronism typing diagnosis system of the present invention

[0029] 1. Research subjects

[0030] This study is a retrospective study, and the subjects included patients who were diagnosed with PA and successfully completed AVS in the Department of Endocrinology and Metabolism of West China Hospital, Sichuan University from January 2018 to December 2022. This invention was reviewed by the Ethics Committee of West China Hospital, Sichuan University (Ethics No.: 2018 Review (No. 297)), and all enrolled patients signed informed consent.

[0031] 2. Research Methods

[0032] All patients were treated according to the expert consensus on the diagnosis and treatment of primary aldosteronism by the Endocrinology Branch of the Chinese Medical Association. [5] And our hospital screening criteria [6]Diagnosis. Exclude factors that affect aldosterone and renin (such as antihypertensive drugs, hypokalemia, etc.). If the aldosterone / renin ratio (ARR) is ≥2.4 (ng / dL) / (mU / L), it indicates a positive screening result. At least one confirmatory test is performed, and a positive result is confirmed as PA. All confirmed patients undergo adrenal CT or MRI examination. If CT shows abnormal adrenal morphology and the patient is willing to undergo surgery, AVS is performed to determine the dominant side. Typical adenomas defined by imaging meet the following characteristics: unilateral nodule diameter >1cm, round or quasi-round, clear boundaries, low density, and no atrophic changes in the ipsilateral and contralateral adrenal glands of the adenoma. Successful AVS cannulation was defined as selectivity index (SI) ≥2 (non-ACTH stimulation) or ≥3 (ACTH-stimulated AVS); the dominant side was determined by lateralization index (LI) ≥3 (non-ACTH stimulation) or ≥4 (ACTH stimulation), or LI was 2-4 with contralateral adrenal vein aldosterone / cortisol ratio (ACR) suppressed (lower than inferior vena cava ACR). The UPA diagnostic criteria are: successful AVS indicates unilateral aldosterone hypersecretion. The BPA diagnostic criteria are: successful AVS shows no dominant side.

[0033] The general clinical data, routine biochemical indexes, imaging data and hormone test results of the two groups of patients were collected. 24-hour urine electrolytes, 24-hour urine creatinine (24h-UCR) and 24h-UAC were tested. Patients with UPA who underwent unilateral adrenalectomy were followed up for at least 6 months after surgery to observe changes in blood pressure, blood potassium, aldosterone and renin, and the remission was judged according to the PA surgical outcome (PASO) standard. The value of 24h-UAC in the diagnosis of PA typing was evaluated by drawing the receiver operating characteristic (ROC) curve.

[0034] Detection method: Plasma aldosterone concentration (PAC) and plasma renin concentration (PRC) were detected by chemiluminescence method, and the kits were from Diasorin, Italy. The intra-assay variability of PAC was 2.1%-4.2%, the inter-assay variability was 5.8%-10.5%, and the detection range was 3-100 ng / dL; the intra-assay variability of PRC was 2.1%-4.9%, the inter-assay variability was 6.8%-13.0%, the standing reference interval was 4.4-46.1 mU / L, and the detection upper limit was 500 mU / L. 24h-UAC was detected by LC-MS / MS (the instrument was CalQuant-S, made in China). The reagents were from the mass spectrometry laboratory of Hangzhou Dian Company. The reference range was 2-20ug / 24h. The detection process was as follows: (1) Sample pretreatment: The urine sample was mixed with HCl in a 1.2mL 96 plate and acid-hydrolyzed at 37°C; the internal standard was added to the acid-hydrolyzed urine sample and vortexed to mix; all samples were added to the SLE plate and the extraction solution was added twice; all eluents were collected, dried with nitrogen, and reconstituted with methanol and water. (2) LC-MS / MS detection: The analyte was taken and chromatographed using a reverse phase column and a triple quadrupole mass spectrometer (Sciex Triple Quad TM The samples were analyzed by LC-MS / MS in multiple reaction monitoring (MRM) mode using a 4500MD instrument.

[0035] The calibration curve used in the laboratory was used to calibrate the analytes. The peak area of ​​24h-UAC showed a good linear relationship with the mass concentration with a correlation coefficient R ≥ 0.995 (R2 ≥ 0.99) within the linear range under the condition that the integrated signal exceeded the signal-to-noise ratio of 10. The accuracy of the lower limit of measuring interval (LLMI) was 13.24% with a Bias% and a precision CV% of 2.54%. The intra-assay precision was 1.84% to 3.10%, the accuracy was -4.03 to 9.19%, and the inter-assay precision was 2.13% to 5.56%. The analyte concentrations were calculated from the integrated chromatograms using the corresponding response factors determined from the appropriate calibration curves in the matrix.

[0036] 3. Statistical methods

[0037] SPSS25.0 statistical software was used to analyze and process the data. The measurement data that conformed to the normal distribution were expressed as mean ± standard deviation (x ± s), and the measurement data that were not normally distributed were expressed as median and interquartile range. Two independent sample t-tests or non-parametric Mann-Whieity U tests were used for comparison. Count data were expressed as frequency (N) and compared using chi-square test. Medcalc 20.218 statistical software was used to draw the ROC curve, and the optimal cut-off value was determined when the Youden index was the largest. When P < 0.05, the difference was statistically significant.

[0038] 4. Results

[0039] 4.1 Comparison of clinical indicators between patients with unilateral and bilateral PA

[0040] According to the research method, 86 PA patients with successful AVS were finally included, including 51 UPA and 35 BPA. There was no statistical difference in gender, age, BMI, systolic blood pressure, diastolic blood pressure, 24-hour urine potassium, 24-hour urine sodium, 24-hour urine chloride, 24-hour urine potassium, 24h-UCR, and PRC between the UPA and BPA groups, P>0.05. The lowest blood potassium level in the UPA group was significantly lower than that in the BPA group, and PAC, 24h-UAC, ARR, and 24h-UAC / UCR were significantly higher than those in the BPA group, and the differences were statistically significant (P values ​​were all <0.05). There was a significant difference in the imaging manifestations of typical adenoma (yes / no) between the two groups, and the difference was statistically significant (P <0.05). The above results are shown in Table 1. In the UPA group, 35 patients underwent adrenal tumor resection, 20 patients were judged to be relieved by the PASO standard after surgery, and 15 patients were not followed up.

[0041] Table 1: Comparison of clinical indicators between UPA group and BPA group

[0042]

[0043] 4.2 The value of 24h-UAC in PA typing diagnosis and establishment of diagnostic model

[0044] The 24h-UAC of UPA patients was significantly higher than that of BPA patients. Figure 1 When 24h-UAC was used to distinguish unilateral lesions, the AUC was 0.829 (95% CI 0.733-0.902), the optimal cut-off value was 15.4μg / 24h, the sensitivity was 68.63%, the specificity was 88.57%, P < 0.0001, see Figure 2 When the cut-off value was >24.5μg / 24h, the specificity was 100% and the sensitivity was 27.45%.

[0045] Univariate logistic regression analysis showed that 24h-UAC, lowest serum potassium level, ARR, PAC, and imaging manifestations of typical adenoma were significantly different between the UPA group and the BPA group. The results are shown in Table 2.

[0046] Table 2: Univariate Logistic regression analysis of factors affecting PA classification

[0047]

[0048] According to the results of Logistic regression in Table 2, multivariate logistic regression analysis determined 24h-UAC, lowest blood potassium level, and imaging manifestations of typical adenoma as independent variable indicators for typing diagnosis. Multivariate Logistic regression analysis was performed with the two types of PA as dependent variables (UPA=1, BPA=0). The final prediction model for typing diagnosis was: F=2.097+0.166×24h-UAC-1.660×lowest blood potassium level+1.310×imaging manifestations of typical adenoma, see Table 3.

[0049] Table 3: Multivariate Logistic regression analysis of factors affecting PA classification

[0050]

[0051] The diagnostic efficacy of the above prediction model for distinguishing UPA from BPA was further analyzed, with an AUC of 0.889 (95% CI 0.803, 0.947), an optimal cut-off value of 0.704, a sensitivity of 74.51%, and a specificity of 91.43%, P < 0.001, see Figure 3 , Hosmer-lemeshow test P = 0.618 (> 0.05), indicating a high degree of fit. When the model cut-off value is 0.815, the specificity is as high as 100%.

[0052] 4.3 Comparison with published Chinese population typing diagnostic models

[0053] Using the Chinese modified Kupers scoring model [7] When the CONPASS prediction model was used to distinguish whether the patient had unilateral lesions, the AUC was 0.812 (95% CI 0.713-0.888). When the cut-off value was 4 points, the sensitivity was 52.94% and the specificity was 85.71%. When the cut-off value was 5 points, the sensitivity was 29.41% and the specificity was 100%, P < 0.001. [8]When used to distinguish whether the patient of the present invention has unilateral lesions, the AUC is 0.688 (95% CI 0.579-0.784), the sensitivity is 49.02%, the specificity is 88.57%, and P<0.001. The diagnostic efficacy of the three models was compared. The positive predictive value of the model of the present invention is higher than that of the other two models, and the missed diagnosis rate is lower (see Figure 4 , Table 4).

[0054] Table 4: Comparison of evaluation indicators of the model of the present invention, the Chinese improved version of Kupers score and the CONPASS model

[0055]

[0056]

[0057] The present invention uses LC-MS / MS to detect 24h-UAC and finds that the 24h-UAC of UPA patients is significantly higher than that of the BPA group, suggesting that it may have a high value in PA typing diagnosis.

[0058] Further analysis confirmed that 24h-UAC has good distinguishing efficacy in distinguishing unilateral and bilateral lesions. When the AUC was 0.829 (95% CI 0.733-0.902), the optimal cut-off value was 15.4μg / 24h, the sensitivity was 68.63%, and the specificity was 88.57%; when the cut-off value was 24.5μg / 24h, the specificity reached 100%. We further analyzed and found that 24h-UAC, lowest blood potassium level, and imaging manifestations of typical adenoma were independent predictors of PA classification. Finally, the diagnostic efficiency of the prediction model established based on these three clinical indicators was significantly improved, with an AUC of 0.889. When the optimal cut-off value was 0.704, the sensitivity was 74.51%, and the specificity was 91.43%. When the cut-off value was 0.815, the specificity was as high as 100%, which can reliably diagnose UPA, making it possible for this group of patients to avoid AVS and undergo direct surgical treatment. Therefore, it has certain potential value in clinical practice.

[0059] Using clinical indicators to establish a classification prediction model has always been one of the research directions related to PA classification diagnosis. Kupers et al. found through regression analysis that blood potassium <3.5mmol / L, typical adenoma imaging manifestations and eGFR are influencing factors for predicting UPA. Based on this, the Kupers scoring system was established, with a total score of 7 points. When the score is ≥5 points, the specificity of diagnosing UPA is as high as 100%. [9]However, a study found that the sensitivity of this scoring system in Chinese PA patients was reduced to 62%, and the specificity was only 53%; but after incorporating 24h-UAC into the diagnostic model, a modified Kupers scoring system was established that included 24h-UAC, a history of hypokalemia, and CT findings of typical adenomas larger than 1 cm. When the score was 5 points, the specificity of the classification diagnosis of Chinese PA patients was increased to 90.5%, and the sensitivity was 45.3%. [7] In addition, the CONPASS team's research suggests that when the biochemical conditions of PA patients meet PAC ≥ 200pg / ml, PRC ≤ 5μIU / mL, and blood potassium ≤ 3.5mmol / L, and CT shows a unilateral adrenal nodule ≥ 1cm (the contralateral side is normal), they are all UPA. [8] When all patients of the present invention were treated with the above two prediction models to distinguish unilateral lesions and compared with the model of the present invention, the prediction efficiency of the model of the present invention was still high and the missed diagnosis rate was lower.

[0060] The present invention found that the 24h-UAC of patients with unilateral PA was significantly higher than that of patients with bilateral PA, which can be used for the classification diagnosis of PA; the prediction model established by combining the three indicators of 24h-UAC, the lowest blood potassium level, and the imaging manifestation of typical adenoma has a high diagnostic efficacy for the classification diagnosis of PA, which may provide assistance for the clinical classification of PA.

[0061] References:

[0062] [1] Yin Y, Yu S, Qiu L, et al. Establishment of a rapid and simple liquid

[0063] chromatography tandem mass spectrometry method for measuring aldosterone in urine. J Chromatogr B Analyt Technol Biomed Life Sci. 2019; 1113:84-90.

[0064] doi:10.1016 / j.jchromb.2019.03.012

[0065] [2] Kobayashi H,Haketa A,Ueno T,et al.Subtype prediction in primary

[0066] aldosteronism:measurement of circadian variation of adrenocortical

[0067] hormones and 24-h urinary aldosterone.Clin Endocrinol(Oxf).

[0068] 2016;84(6):814-821.doi:10.1111 / cen.12998

[0069] [3] Ma Wenjun, Bian Jin, Lou Ying, et al. Clinical application value of 24-hour urinary aldosterone detection by tandem mass spectrometry in the screening of primary aldosteronism[J]. Chinese Journal of Laboratory Medicine, 2020, 43(3): 261-266.

[0070] DOI:10.3760 / cma.j.issn.1009-9158.2020.03.013[3][3]

[0071] [4] Zhang Qianting, et al., The value of 24-hour urinary aldosterone in the diagnosis and classification of primary aldosteronism, Chinese Medical Journal, Vol. 104, No. 45, December 3, 2024

[0072] [5] Chinese Medical Association Endocrinology Branch. Expert consensus on diagnosis and treatment of primary aldosteronism

[0073] (2020 edition)[J]. Chinese Journal of Endocrinology and Metabolism, 2020, 36(9): 727-736.

[0074] DOI:10.3760 / cma.j.cn311282-20200615-00444.

[0075] [6] Chinese Medical Association Endocrinology Branch. Expert consensus on diagnosis and treatment of primary aldosteronism

[0076] (2020 edition)[J]. Chinese Journal of Endocrinology and Metabolism, 2020, 36(9): 727-736.

[0077] DOI:10.3760 / cma.j.cn311282-20200615-00444.

[0078] [7]Zhang Y,Niu W,Zheng F,et al.Identifying unilateral disease inChinese

[0079] patients with primary aldosteronism by using a modified predictionscore.J

[0080] Hypertens.2017;35(12):2486-2492.doi:10.1097 / HJH.0000000000001488 / / [8] Song Y,Yang J,Shen H,et al.Development and validation of model for

[0081] sparing adrenal venous sampling in diagnosing unilateral primary

[0082] aldosteronism.J Hypertens.2022;40(9):1692-1701.

[0083] doi:10.1097 / HJH.0000000000003197

[0084] [9] Küpers EM,Amar L,Raynaud A,Plouin PF,Steichen O.Aclinicalpredictionscore to diagnose unilateral primary aldosteronism.J ClinEndocrinol Metab.2012;97(10):3530-3537.doi:10.1210 / jc.2012-1917.

Claims

1. A method for constructing a primary aldosteronism typing diagnosis model, characterized in that: It includes the following steps: a. Collect clinical data: 24-hour urine aldosterone, lowest blood potassium level, and imaging findings of typical adenoma as indicators; b. The prediction model established by drawing the receiver operating characteristic curve is used to diagnose the type of primary aldosteronism.

2. The method for constructing a primary aldosteronism typing model according to claim 1, characterized in that: The prediction model formula described in step b is: F = 2.097 + 0.166 × 24-hour urine aldosterone - 1.660 × lowest serum potassium level + 1.310 × imaging findings of typical adenoma; Among them, F is the regression equation logist(P), which is the quantitative relationship between the dependent variable primary aldosteronism typing diagnosis and the three independent variables of 24-hour urine aldosterone, lowest blood potassium level, and imaging manifestation of typical adenoma established by regression analysis; 24h-UAC is the total amount of urinary aldosterone in 24 hours measured by liquid chromatography tandem mass spectrometry, expressed in μg / 24h; The unit of the lowest blood potassium level is mmol / L; If the imaging findings were typical adenoma, yes = 1, no = 0.

3. The method for constructing a primary aldosteronism typing model according to claim 1 or 2, characterized in that: It includes the following steps: a, 24-hour urinary aldosterone measured by liquid chromatography-tandem mass spectrometry; b. The lowest value of blood potassium level detected; c. Evaluate imaging findings; d. Calculate the F value according to the prediction model formula of claim 2; e. Primary aldosteronism is classified according to the F value.

4. A typing model constructed by the method for constructing a typing model for primary aldosteronism according to any one of claims 1 to 3.

5. A method for predicting the typing of primary aldosteronism using the primary aldosteronism typing diagnosis model according to claim 4, characterized in that: The F value is greater than or equal to 0.704, indicating unilateral primary aldosteronism. The F value is less than 0.704, indicating bilateral primary aldosteronism.

Images