Analytical Interference in Chemiluminescence Assay–Measured Angiotensin I, Angiotensin II, Aldosterone, and Renin

ABSTRACT

Background

The interference can be a significant source of laboratory errors with the potential to cause immunoassay results to drift. Therefore, we evaluated the interference in various endogenous and exogenous substances on immunoassay for angiotensin I (Ang I), angiotensin II (Ang II), aldosterone, and renin in vitro.

Methods

Ten endogenous and eight exogenous substances were evaluated at supraphysiologic or supratherapeutic plasma levels using the screening study to identify potential interfering substances. Subsequently, potential interfering substances were further tested within maximum pathological or therapeutic plasma concentration ranges using the dose–response study to determine whether the interference has a significant bias. According to preset acceptance criteria, the interference in potential interfering substances for Ang I, Ang II, and renin and aldosterone assays was determined.

Results

Six potential interfering substances for Ang I immunoassays were identified, namely valsartan, nifedipine, spironolactone, cholesterol, hemoglobin, and triglyceride. Meanwhile, ethanol, nifedipine, spironolactone, heparin sodium, warfarin, hemoglobin, uric acid, cholesterol, and triglyceride appeared to have potential interference in the Ang II assay. Three identified as possible interferents for aldosterone immunoassays were glucose, valsartan, and spironolactone. Moreover, warfarin, valsartan, spironolactone, uric acid, cholesterol, bilirubin unconjugated, triglyceride, and hemoglobin were potential interfering substances for renin immunoassays. However, only spironolactone of these potential interfering substances exceeded preset mean bias limits (less than ±10.0%) in aldosterone immunoassays.

Conclusion

Exogenous spironolactone caused clinically significant interference in aldosterone immunoassays. Moreover, the interference in other substances was acceptable in Ang I, Ang II, and renin and aldosterone immunoassays.

In our previous study, we evaluated the interference in various endogenous and exogenous substances on immunoassay for Ang I, Ang II, aldosterone, and renin in vitro according to CLSI Guideline EP07‐A2. The results showed that exogenous spironolactone caused clinically significant interference in aldosterone immunoassays. Moreover, the interference in other substances was acceptable in Ang I, Ang II, and renin and aldosterone immunoassays.

1. Introduction

Immunoassay remains the method of choice in the clinical laboratory for the analysis of many analytes, particularly complex heterogeneous molecules [1]. Moreover, immunoassays are analytically sensitive but lack adequate specificity and accuracy. The specificity of an immunoassay does not only depend on the binding property of the antibody but also the composition of the antigen and its matrix is important [2]. Meanwhile, the specificity can also be influenced by reagent buffer and immunoassay format [3]. However, immunoassay involves the reaction of complex biological reagents with other complex biological reagents in a variable matrix, and they are inherently vulnerable to different types of interference. Therefore, interference in immunoassays is a serious and underestimated problem [4, 5].

As well known, the renin–angiotensin–aldosterone system (RAAS) plays a key role in blood pressure regulation [6]. Therefore, the accurate measurement results of Ang I, Ang II, aldosterone, and renin are critical to ascertain the status of the RAAS, particularly in pathological conditions or following therapeutic intervention, which has been claimed to be useful not only to diagnose secondary form of hypertension but also to guide antihypertensive therapy in patients [7]. For Ang I, Ang II, aldosterone, and renin immunoassays, the analytical performance in terms of precision, accuracy, and immunity to interference is key to reliable results [8]. However, many studies mainly focused on the precision, linearity, accuracy, and carryover of immunoassay [9, 10]. Meanwhile, we found that the interference assessment was incomplete and not comprehensive.

Of note, the erroneous results with such interference may significantly and adversely affect clinical management. If unrecognized, where results are more likely to lead to the misinterpretation of a patient's results by the laboratory and the wrong course of treatment being given by the physician [11]. Hence, the potential interference source of Ang I, Ang II, and aldosterone and renin measurements should be comprehensively investigated. According to the interference in immunoassays [4, 12], sample abnormality interference sources included hemolysis, lipemia, and icterus [13], which may result in matrix, spectral, or chemical interference [14]. Many studies [15, 16] reported that abnormal biochemical metabolite interference sources caused the potential interfering substances including creatinine, uric acid, cholic acid, glucose, and pH value beyond physiological limits. These potential interfering substances may exhibit physical, chemical, spectral, enzymatic, or matrix interference, and uric acid and pH values beyond physiological limits may exhibit enzymatic activity inhibition [17].

Furthermore, the interference in dietary ingested substances is rarely reported. Usually, among therapeutic drugs for target patients, warfarin and heparin sodium are medications used in the prophylaxis and the treatment of venous thrombosis and thromboembolic events [18]. Metformin is of particular concern as it is a very commonly used hypoglycemic medication [19]. Valsartan, spironolactone, and captopril as targeting RAAS drugs were used in the treatment of hypertension [20, 21]. Nifedipine as a calcium antagonist is also used to treat hypertension [22]. However, it is unclear whether these drugs interfere with a chemiluminescent immunoassay for Ang I, Ang II, aldosterone, and renin assays. Meanwhile, few studies [23, 24] reported the interference in these drugs. In particular, the drugs were used to interfere with the RAAS axis to treat hypertension.

Despite the manufacturer providing common interference assessment including the interference in sample additives such as coagulants, biotin, and antibodies sensitivity, we found that the interference assessment of Ang I, Ang II, aldosterone, and renin assays in chemiluminescent immunoassay was also not comprehensive. On the basis of the interference mechanisms of immunoassays [3], comprehensive interference evaluation is essential to ensure measurement value reliability in immunoassays.

Because laboratory experts need it to determine whether measured values meet clinically acceptable standards for measurement results with interference, the clinical laboratory should evaluate the interference effect for Ang I, Ang II, aldosterone, and renin assays in chemiluminescent immunoassay. Therefore, this study was designed to evaluate the interference in various endogenous and exogenous substances in chemiluminescence assay in vitro. Above all, various endogenous and exogenous substances were evaluated by the screening study to identify potential interfering substances. Subsequently, each potential interfering substance was tested using the dose–response study in vitro, and then to determine whether it was an interfering substance according to preset acceptance criteria.

2. Materials and Methods

2.1. Materials and Samples

All chemistry reagents were high‐performance liquid chromatography grade (purity ≥99.0%). Uric acid, cholic acid, glucose, cholesterol, captopril, nifedipine, heparin sodium, spironolactone, valsartan, ethanol, and triglyceride were purchased from Aladdin (Shanghai Aladdin Biochemical Technology Co., Ltd., Shanghai, China). Sodium hydroxide solution (0.1 M), deionized water, phosphate buffer (PBS, pH 7.0), and hemoglobin were obtained from Lirimax (Lirimax (Tianjin) Medical Technology Co., Ltd., Tianjin, China). Metformin hydrochloride, bilirubin unconjugated, creatinine, bilirubin conjugate, warfarin, dimethyl sulfoxide (DMSO), and valsartan were purchased from Sigma‐Aldrich (Sigma‐Aldrich Inc., St. Louis, MO, USA).

Angiotensin I assay kit (chemiluminescent method, Lot: 092230111), Angiotensin II assay kit (chemiluminescent method, Lot: 093230111), aldosterone assay kit (chemiluminescence immunoassay, Lot: 059220211), and renin assay kit (magnetic particle chemiluminescence method, Lot: 428220211) were purchased from Shenzhen New Industry (Shenzhen New Industry Biomedical Engineering Co., Ltd., Shenzhen, China).

From February 2023 to June 2023, the plasma specimens were obtained from leftover patient samples from the Department of Laboratory, the 909th Hospital of Xiamen University. The plasma samples were centrifuged at 2500 g for 10 min at 4°C, and then, the supernatants were transferred to new tubes and stored at a −70°C refrigerator.

2.2. Instrument

An Automatic Chemiluminescence Analyzer (MAGLUMI X8; Shenzhen New Industry Biomedical Engineering Co., Ltd., China) was utilized in the whole evaluation. The analyzer was calibrated once, at the beginning of the measurement, according to the manufacturer's instructions using manufacturer‐provided calibrators.

2.3. Methods

2.3.1. Performance Verification of Analyzer

Before conducting the interference experiment, the performance including precision, accuracy, and carryover of the Automatic Chemiluminescence Analyzer must be evaluated according to CLSI Guidelines [25, 26]. Meanwhile, we assessed whether the analytical system is in control following CLSI Guideline [27]. Details of the evaluation are available in Supplementary Material S1.

2.3.2. Selection of Substances for Interference Testing

According to the Clinical Laboratory Standards Institute (CLSI) Guideline EP07‐A2 [28] and the interference in immunoassay [16, 29, 30, 31], the potential interference sources were investigated, which come from sample abnormalities, abnormal biochemical metabolites, and therapeutic drugs for target patients, dietary ingested substances, and the others, respectively. The possible interfering substances that come from potential interference sources are described in Table S1.

2.3.3. Interference Screening

The interference screening study was performed using plasma samples spiked with or without endogenous or exogenous substances in vitro. Moreover, the testing concentration of each substance was set according to CLSI Guideline EP07‐A2 [28] recommendation and published previously in literatures [32, 33, 34]. These testing concentrations vary according to the specific substance, but all substances were tested at maximum plasma concentration levels expected in the intended patient population. The endogenous substance levels of matrix plasma samples are shown in Table S2 (because of the large difference between endogenous substance levels of matrix plasma samples and its testing concentration, the matrix of endogenous substances will be ignored). The detailed testing concentrations of different substances are shown in Table 1. On the basis of the set testing concentration, the stock solution (20×) of each substance was prepared. The detailed preparation materials and methods are available in Table S3.

TABLE 1.

Testing concentration of endogenous and exogenous substances in the screening study.

Endogenous substance		Exogenous substance
Substance	Testing concentration	Substance	Testing concentration
Hemoglobin	2.00 g/L	Valsartan	0.46 mmol/L
Creatinine	2.00 mmol/L	Ethanol	86.80 mmol/L
Cholesterol	20.00 mmol/L	Spironolactone	1.44 μmol/L
Triglyceride	20.00 mmol/L	Captopril	23.00 μmol/L
Uric acid	1.00 mmol/L	Nifedipine	1.16 μmol/L
Bilirubin conjugate	342.00 μmol/L	Metformin	310.00 μmol/L
Cholic acid	70.00 μmol/L	Heparin sodium	3000.00 IU/L
pH	8.00	Warfarin	32.50 μmol/L
Glucose	110.00 mmol/L	—	—
Bilirubin unconjugated	342.00 μmol/L	—	—

The number of replicates was calculated with the following formula:

$$n=2{\left[\left(z_{\left(1–\frac{\alpha }{2}\right)}+z_{\left(1–\beta \right)}\right)s/d_{\max }\right]}^2$$ (1)

where $z_{\left(1–\alpha /2\right)}$ is the percentage corresponding to the normal distribution at 100 (1 − α)% confidence level of the two‐sided test, $z_{\left(1–\beta \right)}$ is the percentage corresponding to normal distribution at 100 (1 − β)% confidence level, s is the intrabatch repeatability standard deviation of the evaluated method, and $d_{\max }$ is the interference standard at measured experimental concentration in Equation 1. Calculated d _max/s = 2.51. According to d _max/s = 2.51 and to check Table S4, we confirmed that the number of repeated determinations is five times.

The samples of the screening study were prepared and measured with the following procedure: 38.0 mL plasma sample was divided into two equal portions, 1.0 mL stock solution was added to one portion as a test pool and 1.0 mL solute buffer was added to the other portion as a base pool. Then, the testing samples were mixed well and must be rewarmed up at 20–30°C for 30 min. The 20.0 mL test pool was divided into five portions (each 4.0 mL) used as the testing samples (T₁, T₂, T₃, T₄, and T₅, respectively). Furthermore, 20.0 mL base pool was also divided into five portions (each 4.0 mL) used as control samples (C₁, C₂, C₃, C₄, and C₅, respectively). The testing sequence was C₁‐T₁‐C₂‐T₂‐C₃‐T₃‐C₄‐T₄‐C₅‐T₅, and the number of replicates was five times.

2.3.4. Interference Determining

The dose–response study was performed using the plasma samples spiked with or without different potential interfering substances in vitro. The testing concentrations of each potential interfering substance were set within the maximum pathological or therapeutic plasma concentration ranges expected in the intended patient population.

The samples of the dose–response study were prepared and measured with the following procedure: 19.0 mL plasma sample was divided into two equal portions: 0.5 mL stock solution was added to one portion as a test pool, and the other portion was added into 0.5 mL solute buffer as a base pool. The testing samples were mixed at ratios of 2:0, 1.5:0.5, 1.0:1.0, 0.5:1.5, and 0:2 for a total of five concentrations with test and base pools. Then, five mixed samples were obtained, and potential interfering substance concentrations of each sample were available in Table 2. The samples were measured for three replicates on an automatic chemiluminescence analyzer. The mean bias limits were set to less than ±10.0%.

TABLE 2.

Testing concentrations of potential interfering substances in the dose–response study.

Specimen	Testing concentration
	1	2	3	4	5
Specimens with uric acid (mmol/L)	0.15	0.36	0.57	0.78	1.00
Specimens with triglyceride (mmol/L)	1.00	3.25	5.50	7.75	10.00
Specimens with cholesterol (mmol/L)	3.00	7.25	11.50	15.75	20.00
Specimens with bilirubin conjugate (μmol/L)	2.00	87.00	172.00	257.00	342.00
Specimens with hemoglobin (g/L)	1.00	1.25	1.50	1.75	2.00
Specimens with glucose (mmol/L)	5.00	31.25	57.50	83.75	110.00
Specimens with ethanol (mmol/L)	0.00	21.70	43.40	65.10	86.80
Specimens with spironolactone (μmol/L)	0.00	0.36	0.72	1.08	1.44
Specimens with nifedipine (μmol/L)	0.00	0.29	0.58	0.87	1.16
Specimens with valsartan (mmol/L)	0.00	0.19	0.28	0.37	0.46
Specimens with warfarin (μmol/L)	0.00	8.12	16.25	24.38	32.50
Specimens with heparin sodium (IU/L)	0.00	750.00	1500.00	2250.00	3000.00

2.3.5. Statistical Analysis

The interference screening results were calculated with the following formula:

$$d_{\mathrm{obc}}=\overline{x_{\text{test}}}–\overline{x_{\text{control}}}$$ (2)

$$d_{\mathrm{c}}=\frac{d_{\text{null}}+{\mathit{sz}}_{\left(1-\alpha /2\right)}}{\sqrt{n}}$$ (3)

where $\overline{x_{\text{test}}}$ is the mean value of the test pool; $\overline{x_{\text{control}}}$ is the mean value of the base pool in Equation 2; $d_{\text{null}}$ is the invalid assumed value (usually was 0); s is the intrabatch repeatability standard deviation; $z_{\left(1–\alpha /2\right)}$ is the percentage corresponding to normal distribution at 100 (1 − α)% confidence level; n is the sample repetition time in Equation 3. The results were d _obc ≤ d _c that was considered has no statistically significant.

The interference‐determining results were analyzed using linear regression in the GraphPad Prism version 9.5.1 software (GraphPad, San Diego, CA, USA). The mean bias was calculated from slope and intercept from regression analysis.

3. Results

3.1. Analytical Performance of the Analyzer

The performance of the automatic chemiluminescence analyzer was evaluated. As shown in Table S5, the CVs of precision for Ang I, Ang II, aldosterone, and renin‐measured results were <2.0%. These results showed good reproducibility. Meanwhile, the RDs of accuracy for Ang I, Ang II, aldosterone, and renin‐measured results were less than ±2.0%, which indicated the bias was small. Moreover, the carryover of Ang I, Ang II, aldosterone, and renin was <3.50%. The RDs of quality control for Ang I, Ang II, aldosterone, and renin were less than ±5.0%, which met with Westgard multi‐rule quality control criteria [35].

3.2. Potential Endogenous Interfering Substances

We evaluated 10 endogenous substances for the potential interference in Ang I, Ang II, aldosterone, and renin immunoassays using the screening study. The results are available in Table 3. The interference in cholic acid, pH value beyond physiological limits (pH 8.00), bilirubin conjugate, and creatinine for Ang I, Ang II, renin, and aldosterone immunoassays were not found. Moreover, five were identified as potential interfering substances for the renin assay, namely uric acid, cholesterol, triglyceride, bilirubin unconjugated, and hemoglobin. Glucose can influence the aldosterone assay. Moreover, cholesterol, triglyceride, and hemoglobin interfered with the Ang I assay. Furthermore, uric acid, cholesterol, triglyceride, and hemoglobin appeared to potentially interfere in the Ang II assay.

TABLE 3.

Potential interference in endogenous substances in the screening study.

Substance	Testing concentration	Parameter	Control pool measurement value	Testing pool measurement value	d c	d obc
Uric acid	1.00 mmol/L	Ang I	3.13 ± 0.024 ng/mL	3.11 ± 0.034 ng/mL	0.026	−0.020
		Ang II a	83.31 ± 1.414 pg/mL	80.46 ± 1.213 pg/mL	1.709	−2.850
		Aldosterone	94.92 ± 2.093 pg/mL	93.95 ± 2.048 pg/mL	1.768	−0.964
		Renin a	37.07 ± 0.378 uIU/mL	37.51 ± 0.289 uIU/mL	0.344	0.438
Creatinine	2.00 mmol/L	Ang I	3.06 ± 0.0230 ng/mL	3.09 ± 0.0340 ng/mL	0.039	0.034
		Ang II	79.10 ± 1.204 pg/mL	79.01 ± 0.737 pg/mL	0.826	−0.088
		Aldosterone	97.12 ± 2.100 pg/mL	98.18 ± 0.779 pg/mL	1.397	1.060
		Renin	36.82 ± 0.197 uIU/mL	37.01 ± 0.239 uIU/mL	0.201	0.190
Cholesterol	20.00 mmol/L	Ang I a	4.12 ± 0.040 ng/mL	3.99 ± 0.036 ng/mL	0.067	−0.128
		Ang II a	138.92 ± 1.113 pg/mL	144.49 ± 1.185 pg/mL	2.743	5.570
		Aldosterone	63.31 ± 1.324 pg/mL	64.16 ± 1.348 pg/mL	1.172	0.852
		Renin a	34.48 ± 0.203 uIU/mL	32.95 ± 0.152 uIU/mL	0.724	−1.534
d‐glucose	110.00 mmol/L	Ang I	1.32 ± 0.034 ng/mL	1.37 ± 0.050 ng/mL	0.046	0.042
		Ang II	61.42 ± 5.712 pg/mL	63.49 ± 4.954 pg/mL	4.521	2.070
		Aldosterone a	131.88 ± 1.761 pg/mL	135.36 ± 1.451 pg/mL	2.091	3.486
		Renin	49.36 ± 0.200 uIU/mL	49.21 ± 0.210 uIU/mL	0.183	−0.150
Triglyceride	20.00 mmol/L	Ang I a	4.15 ± 0.036 ng/mL	4.31 ± 0.038 ng/mL	0.082	0.166
		Ang II a	95.73 ± 1.081 pg/mL	103.47 ± 1.631 pg/mL	3.755	7.742
		Aldosterone	101.01 ± 1.413 pg/mL	102.06 ± 1.714 pg/mL	1.385	1.046
		Renin a	48.80 ± 0.492 uIU/mL	47.19 ± 0.614 uIU/mL	0.874	−1.610
Bilirubin unconjugated	342.00 μmol/L	Ang I	3.94 ± 0.033 ng/mL	3.97 ± 0.046 ng/mL	0.035	0.024
		Ang II	109.30 ± 1.401 pg/mL	109.00 ± 7.743 pg/mL	4.600	−0.296
		Aldosterone	68.97 ± 2.215 pg/mL	67.36 ± 5.607 pg/mL	3.601	−1.612
		Renin a	35.92 ± 3.450 uIU/mL	39.47 ± 3.364 uIU/mL	3.258	3.548
Cholic acid	250.00 μmol/L	Ang I	4.47 ± 0.106 ng/mL	4.52 ± 0.124 ng/mL	0.098	0.050
		Ang II	49.82 ± 2.074 pg/mL	51.44 ± 2.543 pg/mL	2.058	1.618
		Aldosterone	105.40 ± 1.815 pg/mL	106.39 ± 2.932 pg/mL	2.067	0.996
		Renin	39.17 ± 0.242 uIU/mL	39.29 ± 0.345 uIU/mL	0.252	0.116
Hemoglobin	2.00 g/L	Ang I a	1.04 ± 0.025 ng/mL	0.94 ± 0.038 ng/mL	0.052	−0.098
		Ang II a	60.71 ± 1.447 pg/mL	58.30 ± 2.085 pg/mL	1.854	−2.408
		Aldosterone	96.35 ± 1.643 pg/mL	96.08 ± 1.161 pg/mL	1.182	−0.278
		Renin a	41.47 ± 0.333 uIU/mL	40.81 ± 0.214  uIU/mL	0.385	−0.666
Bilirubin conjugate	342.00 μmol/L	Ang I	4.52 ± 0.124 ng/mL	4.47 ± 0.106 ng/mL	0.098	0.050
		Ang II	52.51 ± 3.204 pg/mL	51.13 ± 1.595 pg/mL	2.058	1.618
		Aldosterone	103.40 ± 1.415 pg/mL	104.39 ± 1.932 pg/mL	2.167	0.996
		Renin	35.92 ± 2.470 uIU/mL	35.47 ± 2.364 uIU/mL	1.258	1.548
pH	8.00	Ang I	0.51 ± 0.009 ng/mL	0.50 ± 0.027 ng/mL	0.018	−0.013
		Ang II	104.41 ± 2.018 pg/mL	104.21 ± 1.409 pg/mL	1.441	−0.201
		Aldosterone	125.85 ± 2.586 pg/mL	127.33 ± 1.159 pg/mL	1.793	1.486
		Renin	33.79 ± 0.053 uIU/mL	33.67 ± 0.289 uIU/mL	0.180	−0.117

^a Potential interfering substance; d _c, cutoff value; d _obc, difference between the mean of the test and control samples.

3.3. Potential Exogenous Interfering Substances

Eight exogenous substances were assessed to identify the potential interference in Ang I, Ang II, aldosterone, and renin‐measured results by the screening study. Metformin and captopril did not interfere with Ang I, Ang II, renin, and aldosterone immunoassays. Moreover, we found that spironolactone, valsartan, and warfarin interfered with the renin assay. Meanwhile, nifedipine, spironolactone, and valsartan influenced the Ang I assay. As expected, ethanol, nifedipine, spironolactone, warfarin, and heparin sodium interfered with the Ang II assay. Spironolactone and valsartan also appeared to have potential interference in the aldosterone assay. All results are available in Table 4.

TABLE 4.

In the screening study, the interference effects of exogenous substances.

Substance	Testing concentration	Parameter	Control pool measurement value	Testing pool measurement value	d c	d obc
Ethanol	86.80 mmol/L	Ang I	1.91 ± 0.053 ng/mL	1.92 ± 0.055 ng/mL	0.045	0.008
		Ang II a	61.92 ± 2.223 pg/mL	63.77 ± 1.363 pg/mL	1.747	1.848
		Aldosterone	89.83 ± 1.489 pg/mL	88.76 ± 1.311 pg/mL	1.260	−1.068
		Renin	42.60 ± 0.513 uIU/mL	42.81 ± 0.084 uIU/mL	0.318	0.204
Spironolactone	1.44 μmol/L	Ang I a	1.17 ± 0.042 ng/mL	1.24 ± 0.060 ng/mL	0.054	0.070
		Ang II a	63.78 ± 1.380 pg/mL	60.67 ± 1.973 pg/mL	2.012	−3.112
		Aldosterone a	105.23 ± 1.430 pg/mL	115.43 ± 1.256 pg/mL	1.209	3.028
		Renin a	42.38 ± 0.255 uIU/mL	42.77 ± 0.190 uIU/mL	0.261	0.398
Nifedipine	1.16 μmol/L	Ang I a	4.18 ± 0.027 ng/mL	4.40 ± 0.023 ng/mL	0.101	0.214
		Ang II a	89.27 ± 1.166 pg/mL	90.58 ± 1.484 pg/mL	1.259	1.314
		Aldosterone	116.21 ± 1.811 pg/mL	114.58 ± 4.779 pg/mL	3.079	−1.626
		Renin	64.46 ± 0.314 uIU/mL	64.18 ± 0.215 uIU/mL	0.285	−0.284
Valsartan	0.46 mmol/L	Ang I a	1.60 ± 0.018 ng/mL	1.64 ± 0.032 ng/mL	0.029	0.044
		Ang II	70.52 ± 0.837 pg/mL	70.51 ± 0.904 pg/mL	0.720	−0.010
		Aldosterone a	104.24 ± 2.076 pg/mL	106.71 ± 1.863 pg/mL	1.989	2.466
		Renin a	39.71 ± 0.294 uIU/mL	39.42 ± 0.210 uIU/mL	0.250	−0.288
Warfarin	32.50 μmol/L	Ang I	1.10 ± 0.021 ng/mL	1.11 ± 0.027 ng/mL	0.020	0.010
		Ang II a	79.51 ± 1.435 pg/mL	76.73 ± 1.425 pg/mL	1.745	−2.780
		Aldosterone	100.52 ± 1.745 pg/mL	100.55 ± 0.729 pg/mL	1.105	0.026
		Renin a	42.23 ± 0.157 uIU/mL	42.54 ± 0.201 uIU/mL	0.206	0.308
Captopril	23.00 μmol/L	Ang I	2.25 ± 0.041 ng/mL	2.30 ± 0.074 ng/mL	0.053	0.044
		Ang II	67.81 ± 1.408 pg/mL	68.68 ± 1.732 pg/mL	1.366	0.876
		Aldosterone	141.82 ± 2.026 pg/mL	141.39 ± 3.055 pg/mL	2.151	−0.436
		Renin	62.61 ± 0.528 uIU/mL	62.45 ± 0.651 uIU/mL	0.495	−0.156
Metformin	310.00 μmol/L	Ang I	1.17 ± 0.037 ng/mL	1.16 ± 0.022 ng/mL	0.026	−0.013
		Ang II	64.59 ± 1.675 pg/mL	64.79 ± 1.326 pg/mL	1.252	0.193
		Aldosterone	128.72 ± 1.649 pg/mL	127.81 ± 2.810 pg/mL	1.950	−0.913
		Renin	46.89 ± 0.235 uIU/mL	47.01 ± 0.452 uIU/mL	0.303	0.118
Heparin sodium	3000.00 IU/L	Ang I	1.32 ± 0.031 ng/mL	1.26 ± 0.037 ng/mL	0.059	−0.059
		Ang II a	36.99 ± 1.365 pg/mL	17.22 ± 0.517 pg/mL	9.170	−19.764
		Aldosterone	132.04 ± 1.411 pg/mL	130.53 ± 1.600 pg/mL	1.629	−1.510
		Renin	52.28 ± 0.320 uIU/mL	51.94 ± 0.175 uIU/mL	0.365	−0.343

^a Potential interfering substance; d _c, cutoff value; d _obc, difference between the mean of the test and control samples.

3.4. Interfering Substances

The interference in potential interfering substances was further evaluated using the dose–response study. The interference assessment of the Ang I assay showed that valsartan and nifedipine caused measured values increase, which appeared to be a positive correlation, and hemoglobin, spironolactone, cholesterol, and triglyceride led to measured values decrease, which appeared to be a negative correlation. However, the mean biases of these potential interfering substances were less than the mean bias limits (less than ±10%). The results suggested that the interference in the above potential interfering substances has no clinical significance in the Ang I assay.

Furthermore, the interference analysis of the Ang II assay showed that nifedipine, triglyceride, and ethanol caused measurement values to increase, which appeared to be a positive correlation, whereas spironolactone, warfarin, heparin sodium, cholesterol, uric acid, and hemoglobin led to measurement values decrease, which showed the negative correlation. Meanwhile, the mean bias of these potential interfering substances did not exceed the mean bias limits. These results also indicated that the interference in the abovementioned potential interfering substances has no clinical difference in the Ang II assay.

For the interference in the aldosterone assay, the linear regression analysis showed that glucose, valsartan, and spironolactone had a positive correlation; meanwhile, we found that the mean bias of spironolactone exceeded the mean bias limits (less than ±10%). Excepted the interference in spironolactone is clinically significant, the mean bias of other interference was acceptable for aldosterone immunoassays.

Moreover, the interference analysis of the renin assay showed that warfarin, bilirubin unconjugated, cholesterol, and valsartan had a positive correlation; triglyceride, hemoglobin, uric acid, and spironolactone had a negative correlation, whereas the mean biases of these potential interfering substances did not exceed the mean bias limits. These results indicated that the interference in the above potential interfering substances has no clinical difference in the renin assay.

In the dose–response study summary, except the mean bias of spironolactone on the aldosterone assay exceeded the mean bias limits (less than ±10%), the mean biases of the other potential interfering substances on Ang I, Ang II, aldosterone, and renin assays were <10.0%. All results are available in Table 5.

TABLE 5.

Linear regression analysis of different potential interfering substances.

Parameter	Substance	Mean bias (%)	Linear regression equation
Ang I	Valsartan	2.34	Y = 0.0722x + 0.80
	Nifedipine	1.61	Y = 0.0225x + 0.81
	Spironolactone	−5.30	Y = −0.1194x + 1.62
	Cholesterol	−1.93	Y = −0.0028x + 1.66
	Hemoglobin	−5.16	Y = −0.0172x + 0.50
	Triglyceride	−5.30	Y = −0.0085x + 0.87
Ang II	Ethanol	1.15	Y = 0.0112x + 42.15
	Nifedipine	1.51	Y = 1.2461x + 48.40
	Spironolactone	−1.50	Y = −1.2158x + 58.14
	Heparin sodium	−4.57	Y = −0.0015x + 49.15
	Warfarin	−3.85	Y = −0.1236x + 52.15
	Hemoglobin	−5.86	Y = −2.253x + 57.64
	Cholesterol	−0.91	Y = −0.0432x + 54.20
	Uric acid	−1.82	Y = −1.4515x + 45.58
	Triglyceride	2.42	Y = 0.2161x + 48.42
Aldosterone	Glucose	4.06	Y = 0.0472x + 66.84
	Spironolactone a	12.41	Y = 22.654x + 131.4
	Valsartan	2.22	Y = 5.8481x + 68.39
Renin	Warfarin	1.20	Y = 0.0199x + 26.91
	Uric acid	−3.30	Y = −1.2365x + 26.71
	Bilirubin unconjugated	5.00	Y = 0.0060x + 20.64
	Valsartan	1.93	Y = 1.5881x + 21.39
	Cholesterol	0.59	Y = 0.0224x + 43.12
	Triglyceride	−2.23	Y = −0.1563x + 38.42
	Spironolactone	−1.93	Y = −1.8180x + 43.80
	Hemoglobin	−1.08	Y = −0.3837x + 52.84

^a The mean bias exceeded preset mean bias limits (less than ±10.0%).

4. Discussion

In this study, we evaluated the interference in various endogenous and exogenous substances on immunoassay for Ang I, Ang II, aldosterone, and renin. Generally, the endogenous interfering substances mainly come from sample abnormalities and abnormal biochemical metabolite interference sources [15]. Moreover, the exogenous interfering substances come from therapeutic drugs for target patients [36], sample additives [30], dietary ingested substances [37], or other interference sources [38, 39]. Therefore, we investigated the possible interfering substances from potential interference sources for chemiluminescent immunoassay. Ten endogenous and eight exogenous substances selected were available in Table S1. Apart from antibody cross‐reactivity, biotin, and coagulant interference, these substances are relatively comprehensive. Meanwhile, these substances can also represent typical matrix, chemical, physical, enzymatic, and spectral interference mechanisms in Ang I, Ang II, aldosterone, and renin immunoassays, which were also commonly potential interfering substances in clinical testing samples.

Although the common interference is caused by various exogenous and endogenous substances of testing samples, it is unclear whether the interference in these substances influences the clinical determination for measured results. Moreover, the interference mechanisms of these substances in immunoassay are complicated and multiple. The problem that interference may cause falsely elevated, normal, and lower values in clinical patient material is very often ignored or not known [40]. Meanwhile, we found that the interference in exogenous and endogenous substances on Ang I, Ang II, aldosterone, and renin immunoassays is rarely reported. Therefore, we should evaluate the interference in these substances and provide scientific and systematic shreds of evidence to determine the reliability of measured results for clinical laboratory experts.

Before the interference assessment, we evaluated the analytical performance of the Automatic Chemiluminescence Analyzer. The results showed that precision and accuracy are excellent, and minimal carryover. According to Laboratory Medicine Practice guideline [41], The Automatic Chemiluminescence Analyzer exhibited excellent analytical performance, and this analytical system is in control, which can meet with interference testing study.

We evaluated the interference in endogenous substances for Ang I, Ang II, renin, and aldosterone immunoassays. Among cholic acid, bilirubin conjugate, creatinine, and pH value beyond physiological limits (pH 8.00) did not interfere with Ang I, Ang II, renin, and aldosterone immunoassays in the screening study. Even if the above substances of the testing sample are at supraphysiological levels, the measured results are still reliable.

Moreover, endogenous cholesterol, hemoglobin, and triglyceride interfered with Ang I, Ang II, and renin immunoassays in the screening study. Similarly, Kazmierczak [42] reported that cholesterol, hemoglobin, and triglyceride caused potential interference in immunoassay. As well known, endogenous cholesterol, bilirubin unconjugated, hemoglobin, and triglyceride cause interference because of spectral interference when these substances achieve a certain concentration within the specimen. However, bilirubin unconjugated as a common endogenous interfering substance only influenced renin measurements and caused measured values to increase. Therefore, we may presume that bilirubin unconjugated caused not a spectral interference, but other unknown interference. Subsequently, the dose–response studies showed that the interference biases of cholesterol, bilirubin unconjugated, hemoglobin, and triglyceride were less than the preset mean bias limits.

The interference in endogenous uric acid caused Ang II and renin measurement values to decrease, which may be associated with chemical interference. An alternative explanation is that uric acid with high concentration may interact with Ang and II renin proteins because of their instability under acidic conditions. Wang et al. [43] studies showed that uric acid cause excessive activation of the RAAS in vivo. The interference results are inconsistent in vivo and in vitro studies. From this point of view, the study in vitro cannot completely mirror truth values in vivo. Another endogenous substance, glucose interfered with aldosterone immunoassays and caused measured value increase. Similarly, Lin et al. [44] reported that impaired glucose tolerance might lead to aldosterone‐measured values being falsely positive. However, the interference bias of uric acid and glucose was less than the preset mean bias limits in this dose–response study. This interference has no clinical significance. Because the interference mechanism of glucose and uric acid is unclear, the role of glucose in the interference in immunoassays is an area worthy of investigation.

In clinical practice, a clear description of interference from endogenous substances is undoubtedly necessary to ensure measured result reliability. Only when the interference bias achieves a certain level, can it influence measured result determination. On the basis of the preset interference concentration of endogenous substances, the interference in these endogenous substances has no clinical significance. Therefore, it is reasonable to conclude that interference from these endogenous substances need not be considered in Ang I, Ang II, renin, and aldosterone immunoassays.

Exogenous substances come from different typical interference sources. In this screening study, captopril as an angiotensin‐converting enzyme inhibitor causes enzymatic conversion disorder [45]. Meanwhile, it is also a RAAS axis blocker, but the interference in captopril for Ang I, Ang II, aldosterone, and renin immunoassays was not found in the screening study. Furthermore, the interference in metformin as a hypoglycemic drug on Ang I, Ang II, renin, and aldosterone immunoassays was not also observed. The patients do not need to discontinue captopril and metformin before undergoing Ang I, Ang II, and aldosterone and renin assays.

However, valsartan appeared to have potential interference in Ang I, Ang II, and renin immunoassays. Valsartan, an Ang II blocker, can block the Ang II receptors and can cause plasma Ang II levels to increase. An alternative explanation is that the conversion of Ang I to Ang II is inhibited after plasma Ang II level increases, resulting in the Ang I plasma level increase. Moreover, although plasma Ang I levels increased, renin consumption in inducing the activity of angiotensin I (Ang I) was reduced, causing plasma renin levels to increase, which is a cascade reaction, which is a typical chemical interference. The interference biases of valsartan for Ang I, Ang II, and renin assays were less than preset bias limits in the dose–response study.

Furthermore, spironolactone, an aldosterone receptor blocker, competes with aldosterone to bind to the corresponding receptor. We found that spironolactone caused Ang I, Ang II, and renin‐measured values to decrease in immunoassays. In contrast, this led to aldosterone‐measured values increase in the aldosterone assay. Similarly, Honour et al. [46, 47] reported that spironolactone interfered with androstenedione and aldosterone immunoassays because spironolactone is an analog of aldosterone, which may lead to typical physical interference or cross‐reactivity.

According to the regulatory mechanisms of RAAS in vivo, the possible reason is that an aldosterone increase can inhibit the expression of Ang I, Ang II, and renin. This study also showed a similar result. However, the interference bias of spironolactone for the aldosterone assay was more than the preset bias limits in the dose–response study. Spironolactone is an interfering substance for the aldosterone assay.

Although cross‐reactivity of RAAS axis blockers has been assessed previously [48, 49], because of the inherited reaction mechanism of immunoassays, it is difficult to completely avoid cross‐reaction, especially for the low‐activity samples. Because of the interference reasons, the Endocrine Society guidelines suggested that testing for Ang I, Ang II, renin, and aldosterone should be ideally performed off‐drugs that interfere with the RAAS axis [50]. However, this study demonstrates that aldosterone receptor antagonists and angiotensin receptor blockers are important interference factors that have to be considered for the interference in Ang I, Ang II, aldosterone, and renin immunoassays. However, the results of testing samples from patients who applied the angiotensin‐converting enzyme inhibitors are reliable.

Of note, nifedipine, a calcium channel blocker drug, was used to treat hypertension, which can interfere with the Ang I and Ang II immunoassays, and cause measured values to increase. Jia et al. [51] studies showed that nifedipine could inhibit the Ang II in vivo. However, the interference effect of nifedipine on the Ang I measurement has rarely been studied. Although the patients were receiving nifedipine, there was a certain bias in the Ang I and Ang II measurement results, but the biases were less than preset bias limits. Therefore, this interference has no clinical significance.

Warfarin and heparin sodium, as different anticoagulants, were used for antithrombotic treatment. In the screening study, we found that warfarin and heparin sodium caused renin‐measured values to increase; on the contrary, warfarin led to the Ang II measurement values decrease. Many studies have demonstrated the prothrombotic effect of RAAS activation [52]. This study also proved that anticoagulant drugs can interfere with renin and Ang II immunoassays. However, in the dose–response study, the interference bias of warfarin and heparin sodium cannot influence measurement result determination. The mean bias was acceptable according to preset bias limits.

Another exogenous substance from dietary ingested, ethanol, can interfere with the Ang II immunoassays and cause its measurement values to increase in the screening study. However, the dose–response studies showed that the interference bias of ethanol for the Ang II assay was less than the preset bias limits. On the basis of the preset plasma ethanol level, the mean bias of interference has no clinical significance. In clinical laboratories, testing samples come from patients after drinking or with chronic alcoholism was common. Meanwhile, the interference mechanism of ethanol on the Ang II immunoassays is unclear and rarely reported. Hence, the role of ethanol in the interference in the Ang II assay is also an area worthy of investigation.

Although in vitro interference tests, such as those used in this study, were not surrogates for in vivo (animal or clinical) investigations, they remain essential tools for the identification and assessment of potential risks and can help guide future studies. For drug interference assessment, this study only assessed some drugs related to hypertension treatment or common medications daily. In this study, the interfering substances selected were also limited. Next, we will evaluate the other interfering substances and will report them in the future.

In summary, the potential interfering substance for Ang I, Ang II, aldosterone, and renin immunoassays was identified using the screening study in vitro. The screening results of interference were shown in the following: (1) six of 18 substances were potential interfering substances for Ang I immunoassays, namely cholesterol, hemoglobin, triglyceride, valsartan, nifedipine, and spironolactone; (2) cholesterol, triglyceride, hemoglobin, uric acid, ethanol, nifedipine, heparin sodium, warfarin, and valsartan appeared as potential interference on the Ang II immunoassays; (3) glucose, spironolactone, and valsartan were identified as potential interfering substances in aldosterone immunoassays; and (4) hemoglobin, triglyceride, cholesterol, uric acid, bilirubin unconjugated, spironolactone, valsartan, and warfarin were identified as possible interferents on renin immunoassays.

For all potential interfering substances, the interfering substance for Ang I, Ang II, aldosterone, and renin immunoassays was determined by the dose–response study. However, only spironolactone was an interfering substance determined for aldosterone immunoassays. The interference in other potential interfering substances on Ang I, Ang II, aldosterone, and renin immunoassays has no clinical significance.

In this study, we provided a comprehensive interference assessment of various endogenous and exogenous substances on Ang I, Ang II, aldosterone, and renin assays. Moreover, our study can provide scientific and systematic shreds of evidence to determine the reliability of their measured results for clinical laboratory experts.

Author Contributions

Shengqiang Liang and Xiaohua Xu participated in the study design; the analysis of the study samples; and the collection, analysis, and interpretation of the data. Yongzhi Xu and Shengqiang Liang involved in the writing of the report. All authors read and approved the final manuscript.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Table S1.

Table S2.

Table S3.

Table S4.

Table S5.

Supplementary Material S1.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.