Distinct lipid profile in haemolytic anaemia-related gallstones compared with the general gallstone

Abstract

Background

Pigment gallstones are not uncommon among patients with chronic haemolytic anaemia. But their clinical characteristics have not been described in detail and not been directly compared with the general gallstone population.

Methods

Patients at Peking Union Medical College Hospital with haemolytic anaemia and subsequent gallstones from January 2012 to December 2022 were included. Cases were matched (1:2) based on age, sex and location of stones to randomly select non-anaemia patients with gallstones (controls).

Results

Screening 899 cases of gallstones, we finally included 76 cases and 152 controls. Total cholesterol (TC), high-density lipoprotein (HDL), and low-density lipoprotein (LDL) for cases were 3.02 ± 0.98 mmol/L, 0.89 ± 0.30 mmol/L and 1.58 ± 0.70 mmol/L, respectively, significantly lower than those in the control group (all p < 0.001). TC and HDL were both lower than the normal range, but triglyceride and LDL were within the normal range. Multiple stones were significantly more common for cases (n = 59, 78%) than for controls (n = 44, 29%, p < 0.001). The mean diameter of the maximal gallstone was 1.2 ± 0.6 cm and 1.5 ± 1.0 cm for cases and controls (p = 0.120), respectively. Stones in the elderly (p = 0.002 for univariate analysis, and 0.001 for multivariate analysis) and stones in the bile duct (p = 0.005 for univariate analysis, and 0.009 for multivariate analysis) were found to occur in a shorter period after anaemia.

Conclusion

The lipid profile of haemolytic anaemia with gallstones was distinct, low TC, low HDL, and increased-to-normal LDL, compared with the general gallstone population. Patients with haemolytic anaemia were recommended an abdominal ultrasound if aged older than 50 years, with more frequent follow-up visits.

KEY MESSAGES

1. Clinical characteristics of gallstones following chronic haemolytic anaemia were described and compared with the general gallstone population.

2. The lipid profiles were distinctly different between the patients with gallstones following chronic haemolytic anaemia and the general gallstone population.

3. Elder patients were complicated with gallstones in a shorter period after anaemia and thus were recommended an abdominal ultrasound if aged older than 50 years, with more frequent follow-up visits.

Keywords:

• Gallstones
• haemolytic anaemia
• hypocholesterolemia
• hypolipidaemia

Introduction

Gallstones are prevalent worldwide and mainly of cholesterol type. Typical risk factors include age, female sex and obesity [1]. Acute diseases and complications caused by gallstones, such as acute cholecystitis, acute cholangitis and biliary pancreatitis, constitute the most common emergencies [2]. Cholesterol gallstones are commonly reported [3]. Another type of gallstone, pigment stones, are reported less frequently but are highly prevalent in patients with chronic haemolytic anaemia [4]. Serum level of unconjugated bilirubin from haemolysis increases, with extremely high levels noted in the gallbladder and bile duct, ultimately developing sludge, gravels, and/or stones [5]. However, pigment stones are usually neglected and often misdiagnosed as cholesterol ones in clinical practice [1]. They are however pathophysiologically distinct and are characterized by different primary concerns. Generally, the location and not the composition determined clinical management [6], which could result in overlooking of the primary haemolytic anaemia by both primary physicians and patients. The clinical features of gallstones following haemolytic anaemia have not been described in detail. Gallstones generally are accompanied with hypercholesterolemia [7,8] whereas patients with chronic haemolytic anaemia often had a low level of serum lipids [9]. The diagnosis of hereditary spherocytosis (HS), a type of chronic haemolytic anaemia, was made based on the manifestation of hypercholesterolemia [10]. The lipid profile in these patients was not fully investigated. By study thus aimed to describe the clinical features of patients with haemolytic anaemia complicated with gallstones and compared them with those of the general population with gallstones.

Methods

The STROBE guideline was followed [11]. The institutional review board of Peking Union Medical College Hospital (PUMCH) approved this retrospective study and waived the requirement for written informed consent.

Patients

Patients with a final diagnosis of gallstones and haemolytic anaemia hospitalized from January 2012 to December 2022 at PUMCH were included. Gallstones referred to the presence of stones in the gallbladder, bile duct or both, including cholecystolithiasis and choledocholithiasis, regardless of symptoms and complications. Immune and non-immune-related haemolytic anaemia cases were included, particularly autoimmune haemolytic anaemia (AIHA) and HS. Patients who developed gallstones after the diagnosis of haemolytic anaemia were included. Patients who developed gallstones before the diagnosis of haemolytic anaemia and those who were not in good condition overall, such as those with multiple organ dysfunction, were excluded. Cases were matched (1:2) based on age, sex and location of stones to randomly select non-anaemia patients with gallstones (controls).

Diagnosis

Gallstones were confirmed by transabdominal ultrasound when foci presented strong echoes in the gallbladder, bile duct or both, with acoustic shadowing and gravitational dependency. Anaemia was defined as pre-treatment haemoglobin level lower than 120 g/L for males or 110 g/L for females based on two consecutive tests at least 24 h apart. Haemolytic anaemia referred to anaemia accompanied by reticulocyte percentage higher than 2, percentage of indirect bilirubin (unconjugated bilirubin) higher than 0.5, and any relevant additional test to find a specific cause. The severity of anaemia was classified based on haemoglobin level: 90–110/120 g/L as mild, 60–90 g/L as moderate, 30–60 g/L as severe and <30 g/L as very severe. The additional tests involved peripheral blood smears demonstrating schistocytes or spherocytes, newly positive direct antiglobulin tests (Coombs), and haemoglobin analysis. The test of serum lipids included triglyceride (TG), total cholesterol (TC), high-density lipoprotein (HDL), and low-density lipoprotein (LDL). Hypolipidemia was defined as TC < 3.1 mmol/L or LDL < 1.3 mmol/L [12,13]. Patients with mild HS had normal haemoglobin levels and were included if they fulfilled other criteria.

Data extraction

Electronic medical records were retrospectively reviewed to collect patients’ baseline information, data regarding laboratory tests and clinical courses. The baseline information included age, sex, height, weight, blood pressure, history of hypertension or diabetes mellitus (DM) and length of hospital stay. The laboratory tests included complete blood counts (haemoglobin, platelets, neutrophils and reticulocyte percentage), liver biochemistry tests (total bilirubin, direct bilirubin, alanine transaminase (ALT), aspartate aminotransferase (AST), lactic dehydrogenase (LDH) and γ-glutamyl transferase (GGT)), kidney function tests (creatine (SCR), uric acid (UA) and urea (BUN)), fasting blood glucose and serum lipids. We extracted values obtained from the last two tests before any treatment related to anaemia (for the case group) or gallstones (for the control group), and presented the mean value. The clinical courses generally included features of haematolytic anaemia (onset, causes, treatments and complications) and characteristics of gallstones (onset, symptoms, ultrasound results, acute diseases, treatments and complications). Symptoms related to haemolytic anaemia included fatigue, weakness, shortness of breath, jaundice, dark urine and splenomegaly. Symptoms of acute gallstone diseases included abdominal pain, fever and jaundice. Data were censored on December 2022.

Statistical analysis

Descriptive and continuous variables were presented as numbers with percentages and mean with standard deviation, respectively. The time of gallstone detection was calculated as years from the onset of anaemia to the detection of gallstones. Fisher’s exact test was used for comparing the descriptive variables, and the t-test or Wilcoxon test was used for continuous variables. Statistical analysis was performed using R (4.2.1) for macOS or SPSS 17.0 (IBM Corp., Armonk/NY, USA). A double-sided p < 0.05 was considered to indicate a statistically significant difference.

Results

Demographic characteristics

A total of 899 cases of gallstones were screened, including 106 anaemia cases and 793 non-anaemia controls. After excluding by criteria mentioned in the methods and matching by age, sex, and stone location, 76 cases and 152 controls were included in the analysis. The demographic characteristics are presented in Table 1. There were 37 and 79 males in the case group and the control group, respectively (p = 0.675). Age at gallstone onset was 39 ± 20 years and 41 ± 15 years for the case group and the control group, respectively (p = 0.340). Hypertension was noted in 14 and 32 patients in the case group and the control group, respectively (p = 0.728); 7 and 15 patients in the case and control group had a history of diabetes (p = 1.000), respectively.

Table 1. Baseline characteristics of cases (n = 76) and controls (n = 152).

Characteristics	Cases (n = 76)	Controls (n = 152)	p Value
Age of gallstone onset	39 ± 20	41 ± 15	0.340
Sex, female	39 (51%)	73 (48%)	0.675
Comorbidities
Hypertension	14	32	0.728
Diabetes	7	15	1.000
Blood counts
Haemoglobin (g/L)	84 ± 23	145 ± 17	<0.001
Reticulocyte (%)	11.7 ± 7.5	NA	NA
Neutrophils, ×10^9/L	4.13 ± 1.74	4.07 ± 1.60	0.786
Platelets, ×10^9/L	203 ± 71	240 ± 57	<0.001
Liver biochemical tests
TBIL (μmol/L)	81.04 ± 52.95	14.10 ± 5.48	<0.001
DBIL (μmol/L)	12.69 ± 6.28	4.32 ± 1.75	<0.001
GGT (U/L)	26.22 ± 16.73	43.56 ± 43.25	0.001
LDH (U/L)	260.5 ± 116.1	170.8 ± 31.5	<0.001
ALP (U/L)	82.54 ± 57.76	78.42 ± 27.18	0.139
AST (U/L)	24.92 ± 18.87	26.61 ± 22.92	0.439
BUN (mmol/L)	8.71 ± 27.40	4.80 ± 1.28	0.214
UA (μmol/L)	378.2 ± 114.7	361.5 ± 86.9	0.347
ALT (U/L)	21.57 ± 21.14	30.98 ± 32.17	<0.001
SCR (μmol/L)	62.52 ± 21.35	72.51 ± 15.08	0.001
Lipid profiles
TC (mmol/L)	3.02 ± 0.98	4.59 ± 0.91	<0.001
TG (mmol/L)	1.26 ± 0.65	1.64 ± 1.28	0.318
HDL (mmol/L)	0.89 ± 0.30	1.18 ± 0.24	<0.001
LDL (mmol/L)	1.58 ± 0.70	2.72 ± 0.83	<0.001
Gallstones
Location			0.105
gallbladder	61 (80%)	135 (89%)
biliary tract or both	15 (20%)	17 (11%)
Number
One	12 (16%)	108 (71%)	<0.001
Multiple	59 (78%)	44 (29%)
NA	5 (7%)	0 (0%)
Maximal (cm)	1.2 ± 0.6	1.5 ± 1.0	0.120

ALP: alkaline phosphatase; ALT: alanine transaminase; AST: aspartate aminotransferase; BMI: body mass index; BUN: blood urea nitrogen; DBIL: direct bilirubin; GGT: γ-glutamyl transferase; HDL: high density lipoprotein; LDH: lactate dehydrogenase; LDL: low density lipoprotein; NA: not available; SCR: serum creatinine; TBIL: total bilirubin; TC: total cholesterol; TG: triglyceride; UA: uric acid.

Laboratory tests

Values for haemoglobin, platelets, total bilirubin, direct bilirubin, GGT, LDH, ALT and SCR were significantly different between the case group and the control group (Table 1). However, neutrophil count and levels of ALP, AST, BUN and UA were not significantly different.

The lipid profiles were distinctly different between the case group and the control group. TC, HDL and LDL were 3.02 ± 0.98, 0.89 ± 0.30 and 1.58 ± 0.70 mmol/L, respectively, and were significantly lower in the case group than in the control group (all p < 0.001); TG level was not significantly different (p = 0.319). TC and HDL were both lower than the normal range, but TG level and LDL were within the normal range. Baseline test results are shown in Table 1.

Gallstones

Stones were mainly detected in the gallbladder (n = 61 for the case group and 135 for the control group), some in the bile duct (n = 6 for the case group and 14 for the control group), and a few in both (n = 9 for the case group and 3 for the control group). The number of detected stones was significantly different between the case group and the control group (p < 0.001): most patients in the case group (n = 59, 78%) had multiple stones but most in the control group had only one detectable stone (n = 108, 71%). The mean diameter of the maximal gallstone was 1.2 ± 0.6 cm and 1.5 ± 1.0 cm for the case group and the control group (p = 0.120), respectively.

Patients in the case group were diagnosed with gallstones after 9 (0–52) years of anaemia diagnosis. Thirty-eight patients had acute cholecystitis pertinent to gallstones, including five cases complicated by cholangitis and six by acute pancreatitis. Total bilirubin was 81.0 ± 52.9 μmol/L and 463.7 ± 383.3 μmol/L in the absence and presence of acute gallstone-related diseases, respectively (p < 0.001). Direct bilirubin was 12.7 ± 6.3 μmol/L and 311.0 ± 278.6 μmol/L, respectively (p < 0.001). The ratio of direct/total bilirubin was 0.21 ± 0.12 and 0.60 ± 0.22 for asymptomatic cases and acute cases (p < 0.001), respectively. GGT and LDH were also significantly higher in acute cases than in asymptomatic cases (p < 0.001 and p = 0.02, respectively). Other parameters were not significantly different. Seventy-seven patients in the control group experienced acute diseases. The ratio of direct/total bilirubin for acute cases was not significantly different from that of the case group (p = 0.827), whereas the level of total bilirubin was significantly lower than that of the case group (p < 0.001).

For the case group, stones in the bile duct mostly were subjected to endoscopic retrograde cholangiopancreatography (10/15) to relieve symptoms. Stones in the gallbladder were either followed up via observation (n = 39) or subjected to open or laparoscopic cholecystectomy (n = 31). Eighteen patients had both cholecystectomy and splenectomy in the same operation. For the control group, most patients were recommended observation instead of operation (n = 134). The remaining 18 patients underwent laparoscopic cholecystectomy.

Haemolytic anaemia

The causes of haemolytic anaemia included HS (n = 44), AIHA (n = 22), thalassemia (n = 8) and cold agglutinin disease (n = 2). They were mainly chronic haemolytic anaemia (n = 74). Only two cases of acute haemolysis were reported. The majority of the cases were of moderate anaemia (n = 35), followed by mild (n = 27), and severe (n = 11); no case of very severe anaemia was reported. Three cases of normal haemoglobin were reported. The pre-treatment values for haemoglobin, reticulocyte percentage, neutrophil count and platelet count were 84 ± 23 g/L, 11.7 ± 7.5%, 4.13 ± 1.74 × 10⁹/L, and 203 ± 71 × 10⁹/L, respectively. The age at onset of symptoms related to haemolytic anaemia was 27 ± 21 years. Fifty cases of splenomegaly were reported. The majority of patients in the case group (n = 54) did not have a history of an inherited disorder. Seventeen and five patients in the case group had confirmed or suspected HS and thalassemia, respectively. Most HS patients (n = 23) underwent open or laparoscopic splenectomy. Other HS patients were managed by observation and treatment for symptomatic relief. All AIHA patients received glucocorticoids. Other patients received intravenous immunoglobulin (n = 5), rituximab (n = 5), cyclosporin A (n = 4) and tacrolimus (n = 2). Three patients had a splenectomy. The results related to anaemia are shown in Table 2.

Table 2. Characteristics related to anaemia of the case cohort (n = 76).

Characteristics	Number
Hospital stays (days)	16 ± 13
Hemoglobinopathies
HS	44 (58%)
AIHA	22 (29%)
Thalassemia	8 (11%)
CAD	2 (3%)
Severity of anaemia
Normal	3 (4%)
Mild	27 (36%)
Moderate	35 (46%)
Severe	11 (14%)
Onset of anaemia onset	27 ± 21
Splenomegaly	50 (66%)
Blood counts
Haemoglobin (g/L)	84 ± 23
Reticulocyte (%)	11.7 ± 7.5
Neutrophils (×10^9/L)	4.13 ± 1.74
Platelets (×10^9/L)	203 ± 71
Detection after anaemia of gallstones (years)	12 ± 12

AIHA: auto immune haemolytic anaemia; CAD: cold agglutinin disease; HS: hereditary spherocytosis.

HS versus AIHA

Features of anaemia and gallstones were compared between HS and AIHA (Table 3). HS patients were admitted at a younger age than AIHA patients (32.6 vs. 53.0 years, p < 0.001) but had less severe anaemia before any treatment (haemoglobin, 92 vs. 65 g/L, p < 0.001). The number of female patients with AIHA was significantly higher than that of HS (17/22 vs. 19/44, p = 0.01). Hypolipidemia was more significant in HS than in AIHA (TC, 2.65 vs. 3.71 mmol/L, p = 0.002; LDL, 1.30 vs. 2.10 mmol/L, p = 0.002). Total bilirubin in HS patients was nearly twice as that in AIHA patient (103 vs. 57 μmol/L, p < 0.001). The ratio of direct/total bilirubin was significantly lower in HS than in AIHA (0.16 vs. 0.25, p = 0.008), whereas the ratio was similar for cases of acute disease (0.59 vs. 0.53, p = 0.632). LDH was significantly higher in AIHA than in HS (325 vs. 236 U/L). Other parameters were not significantly different (Table 3). Features of gallstones were not significantly different between HS and AIHA.

Table 3. Comparison between HS and AIHA.

Features	HS (n = 44)	AIHA (n = 22)	p Value
Sex, female	19	17	0.010
Age at admission (years)	33 ± 18	53 ± 17	<0.001
Admission (days)	17 ± 12	20 ± 14	0.316
Blood counts
Haemoglobin (g/L)	92 ± 19	65 ± 22	<0.001
Reticulocyte (%)	11.4 ± 5.9	13.3 ± 10.0	0.423
Platelets (×10^9/L)	202 ± 76	201 ± 73	0.949
Neutrophils (×10^9/L)	3.99 ± 1.47	4.55 ± 2.13	0.271
Liver biochemical tests
TBIL (μmol/L)	103.3 ± 48.9	56.8 ± 47.22	<0.001
DBIL (μmol/L)	14.1 ± 6.9	10.9 ± 4.0	0.020
GGT (U/L)	24 ± 18	31 ± 14	0.129
LDH (U/L)	236 ± 62	325 ± 165	0.025
ALP (U/L)	81 ± 52	71 ± 27	0.333
AST (U/L)	24 ± 21	28 ± 18	0.441
BUN (mmol/L)	5.13 ± 2.10	17.57 ± 49.82	0.255
UA (μmol/L)	400 ± 116	346 ± 124	0.113
ALT (U/L)	18 ± 16	29 ± 30	0.111
SCR (μmol/L)	59 ± 18	68 ± 29	0.182
Lipid profiles
TC (mmol/L)	2.65 ± 0.78	3.71 ± 1.07	0.002
TG (mmol/L)	1.16 ± 0.44	1.53 ± 0.98	0.164
HDL (mmol/L)	0.84 ± 0.30	0.93 ± 0.30	0.353
LDL (mmol/L)	1.30 ± 0.51	2.10 ± 0.82	0.002
Gallstones
Detection after anaemia (years)	14 ± 13	8 ± 8	0.056
Location			0.573
Gallbladder	35	20
Biliary tract	3	0
Both	6	2
Number			1.000
One	5	3
Multiple	35	18
NA	4	1
Maximal (cm)	1.1 ± 0.4	1.1 ± 0.4	0.852

AIHA: auto immune haemolytic anaemia; ALP: alkaline phosphatase; ALT: alanine transaminase; AST: aspartate aminotransferase; BMI: body mass index; BUN: blood urea nitrogen; CAD: cold agglutinin disease; DBIL: direct bilirubin; GGT: γ-glutamyl transferase; HDL: high density lipoprotein; HS: hereditary spherocytosis; LDH: lactate dehydrogenase; LDL: low density lipoprotein; NA: not available; SCR: serum creatinine; TBIL: total bilirubin; TC: total cholesterol; TG: triglyceride; UA: uric acid.

Time to detect gallstone after anaemia

For the case group specifically, univariate analysis showed that the time of gallstone detection (after anaemia) was significantly correlated with age at anaemia onset (p = 0.002) and location of gallstone (p = 0.005): stones in the elderly and stones in the bile duct were found in a shorter period. Multivariate analysis also revealed these two variables to be significant (p = 0.001 and 0.009, respectively). Gallstones were detected in 69% of patients older than 50 years within 5 years of anaemia onset, whereas they were detected after more than 5 years in 76% of patients younger than 20 years, with 55% of patients being diagnosed after more than 10 years. Sex, cause of haemolytic anaemia, the severity of anaemia, lipid profiles and other clinical features were not significantly associated with the time of gallstone detection.

Discussion

To the best of our knowledge, this study was the first to summarize the clinical features of hospitalized patients with gallstones secondary to haemolytic anaemia compared with the general gallstone. We found that these patients had a distinct lipid profile from the general population with gallstones.

A total of 76 cases matched with 152 controls were finally included. The number of cases was comparable to previous studies that reported no more than 60 patients with either HS, or AIHA, or thalassemia [14–18]. We found that anaemia gallstone cases had significantly lower TC, HDL, and LDL than general gallstone controls. As for TC and HDL, hypocholesterolemia and decreased HDL were previously reported in a group composed of both thalassemia and HS cases [6,19]. This was expected in HS due to the defect in red cell membrane proteins and loss of membrane cohesion [20]. Therefore, the profile of low TC and HDL was likely characteristic of hemoglobinopathies. Our further investigation of the subgroups of HS and AIHA demonstrated that such a unique lipid profile was attributed to HS and not correlated with haemoglobin. LDL in the case group remained in the normal range despite being significantly lower than that of the control group. The value of LDL for the case group was also higher than that noted in previous reports. Higher LDL increased the risk of gallstone-related diseases [21,22]. Therefore, an increase in LDL to normal potentially increased the risk of gallstones in the case group.

The case group had an equal number of male and female patients. No female pre-dominance was noted unlike the general population with gallstones (535/793, 67% female in the general gallstone population before screening, data not shown); it was previously reported that female patients were at a higher risk of gallstones than male patients [23,24]. Therefore, the equal number of male and female in the case group was a feature different from the general gallstone group. Another interesting feature was that multiple stones were detected in the case group but only a single stone was reported in the control group. Pigment stones are formed by the accumulation of an excessive amount of unconjugated bilirubin concentrates in the bile, which precipitates in the form of calcium bilirubin and then forms black pigment stones [5,25]. The chronic nature could lead to multiple emergencies, partly explaining the observation of multiple stones in the case group.

We found that the time to detect gallstones was related to the age at anaemia onset and the stone location. Previous reports indicated that age was an independent risk factor for gallstone-related diseases, but did not make a distinction between cholesterol and pigment stones [1,24,26,27]. The results of our study could, to some extent, indicate that age was also a risk factor for pigment gallstones. Therefore, patients with haemolytic anaemia were recommended an abdominal US if aged older than 50 years, with more frequent follow-up visits. The other independent risk factor was the location of the stones. Stones in the bile duct caused more noticeable and obvious symptoms than those in the gallbladder and hence could be more easily detected. However, our results did not reveal a significant association between female sex, anaemic diseases, or severity of anaemia and the time of detection.

We compared the gallstones between HS and AIHA. They were characterized by their primary haemolytic anaemic diseases. HS patients were younger and had less severe anaemia and hypolipidemia and more obvious hyperbilirubinemia. AIHA, as an autoimmune disease, had a female pre-dominance. However, they showed similar clinical features as acute gallstone diseases.

This study has some limitations such as its retrospective nature. Black pigment stones typically are associated with chronic haemolysis. However, we could not re-evaluate the composition of the stones and confirm them as pigment stones resulting from haemolytic anaemia. We did not compare the case group with anaemia patients not having gallstones. Thus, we could not determine the predictors of gallstone occurrence.

Conclusions

The patients with gallstones following chronic haemolytic anaemia had significant lipid characteristics, including low TC, low HDL, but elevated LDL to normal levels. Therefore, LDL was potentially correlated to gallstones and likely involved in the pathogenesis of gallstones. Stones in the elderly and stones in the bile duct were found in a shorter period after haemolytic anaemia. Therefore, patients with haemolytic anaemia were recommended an abdominal ultrasound if aged older than 50 years, with more frequent follow-up visits.

Ethical approval and consent to participate

This study was approved by the ethics committee of Peking Union Medical College Hospital.

Author contributions

Z.W., X.B. and C.H. performed the database search, study selection, data collection, statistical analyses and wrote the manuscript. Y.Z., Y.W., T.S., K.S., M.L., Q.W. and D.W. participated in designing this study and revising the draft. Y.F. and A.Y. supervised this study and revised the draft. All authors approved the final submitted version.

Acknowledgements

No acknowledgement needs to declare here.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The data underlying this article will be shared on reasonable request to the corresponding author.

Additional information

Funding

This work was supported by the CAMS Innovation Fund for Medical Sciences [2021-I2M-1-013], National High Level Hospital Clinical Research Funding and Fundamental Research Funds [2022-PUMCH-A-074 and 2022-PUMCH-A-177], National Key Clinical Specialist Construction Project [ZK108000] and National Natural Science Foundation of China, Joint Fund Project [Integrated Project Grant No. U20A6001]. This funding source had no role in the study design or execution, analyses, interpretation of the data, or decision to submit results.