DOI: https://doi.org/10.21203/rs.3.rs-283746/v1
Background To explore the value of TERT mutations in predicting the early recurrence and prognosis of hepatitis B-related hepatocellular carcinoma (HCC) patients underwent curative hepatectomy.
Methods A total of 81 patients with hepatitis B-related HCC were enrolled and all patients underwent curative hepatectomy. Associations were sought between TERT mutations and recurrence rate within 2 years after hepatectomy, time to progress (TTP) and overall survival (OS).
Results TERT mutations (HR: 2.985, 95%CI: 1.158-7.692, p=0.024) and Barcelona clinic liver (BCLC) stage B (HR: 3.326, 95%CI: 1.019-10.856, p=0.046) were independent risk factors for recurrence within 2 years after hepatectomy. Patients with a TERT mutation had poor TTP (p=0.003) and OS (p=0.013) than others. TERT mutations (HR: 2.245, 95%CI: 1.185-4.252, p=0.013) and BCLC stage B (HR: 2.132, 95%CI: 1.082-4.198, p=0.029) were independent risk factors for poor TTP after curative hepatectomy. A predictive model based on TERT mutations and BCLC stage had better ability to predict early recurrence after hepatectomy of HCC patients than any single factor (AUC: 0.688 vs. 0.639, 0.688 vs. 0.607, respectively). Patients with both TERT mutations and BCLC stage B had poorer TTP and OS than others (p=0.001, p<0.001, respectively).
Conclusion TERT mutations had ability to predict early recurrence and poor prognosis for hepatitis B-related HCC patients underwent curative hepatectomy.
Hepatocellular carcinoma (HCC) ranks sixth in terms of incidence among cancers and fourth in terms of cancer related death [1]. Risk factors of HCC included hepatitis B virus (HBV) or hepatitis C virus (HCV) infection, alcoholism and Aflatoxin intake [2, 3], in which HBV infection is the leading risk factor for HCC [3–5]. Although treatments for HCC are more diversified, curative hepatectomy is still the first choice if the state of patients allowed. However, the recurrence after curative treatment is common in HCC patients [6, 7], which caused a negative effect on survival. Hepatocarcinogenesis is a complex multiple process that involves many genetic alterations [8]. Mutations of tumor suppressor genes or oncogenes promote hepatocarcinogenesis and contribute to post-operative recurrence [9]. A study reported that gene mutations correlated with recurrence in a short term after live transplantation [10]. There is a need to find genes associated with prognosis of HCC patients underwent curative hepatectomy.
Key genes mutated in HCC tissues were identified by Next-generation sequencing (NGS). The most common mutated genes included telomerase reverse transcriptase (TERT), TP53 and CTNNB1 [11–14]. As a catalytic component of telomerase complex, TERT could active the telomerase, thereby causing the telomere lengths maintained and cell survival [15]. TERT is suppressed in most cells of adults [16]. Overexpression of TERT gene increases the activity of telomerase and then leads to telomere synthesis that avoids the senescence and apoptosis caused by telomere shortening [17, 18]. Several studies have uncovered that TERT mutations predicted poor prognosis in solid tumors such as breast cancer and lung cancer [19]. But the correlation between TERT mutations and prognosis of HCC patients was still unclear. In this study, we detected TERT mutations of HCC tissues by NGS and explore the potential prognostic utility of TERT mutations in hepatitis B-related HCC patients underwent curative hepatectomy.
2.1 Patients enrollment
From August 2015 to August 2018, patients underwent curative hepatectomy in the Affiliated hospital of Qingdao University were enrolled based on the following criterion: no macrovascular invasion; no extrahepatic metastasis; with positive serum hepatitis B surface antigen (HBsAg); the pathological diagnosis of tumor was HCC; the surgical margins of specimens were negative; no recurrence in the first 2 months after hepatectomy; abnormal preoperative serum alpha-fetoprotein (AFP) dropped to normal level in the first 2 months after hepatectomy; patients agreed with genetic testing of tumor tissues. All patients signed a written informed consent. The study was conducted in strict compliance with the Helsinki Declaration and was approved by the Ethics Committee (ethics approval number: QDFYKYLLL-20161212). Corresponding epidemiological information and clinicopathological characteristics were collected from the Affiliated Hospital of Qingdao University.
2.2 Identification of genetic alterations and TMB
Formalin-fixed, paraffin-embedded tissues were collected from patients for deep sequencing. The genes were captured and sequenced by genomic profile produced using the NGS-based YuanSu 450 gene panel. Genetic alterations were identified as follows: single nucleotide variants (SNVs) were identified by MuTect (v1.7); Insertion-deletions (InDels) were identified by using PINDEL (V0.2.5). The functional impact of genetic alterations was annotated by SnpEff3.0. Copy number variations (CNV) regions were identified by Control-FREEC (v9.7). Gene rearrangement/fusion were detected through an in-house developed pipeline. Tumor mutational burden (TMB) was measured by an algorithm developed in-house[20].
2.3 Follow up
All patients enrolled in this study were followed up regularly after surgery. During the first 3 months after liver resection, the patients were followed up once a month; during 3-24 months after liver resection, they were followed up every 3 months; and after 2 years, they were followed up every 6 months. The follow up examination included serum AFP, liver function, ultrasonic examination of liver and computed tomography (CT) of lung. Patients received the contrast enhanced CT of upper abdomen annually. When suspected signs of recurrence were found, contrast enhanced CT or magnetic resonance imaging (MRI) was performed to clarify the diagnosis. Time to progress (TTP) was confirmed by imaging examination. All patients were followed up until August 31 2020 or died[20].
2.4 Statistical analysis
All statistical analyses were performed using Statistical Package for the Social Science version 25 for windows (IBM Corp, Armonk, NY). Chi-square test or Fisher’s exact test were used to analyze the association between different parameters and TERT mutations. Chi-square test or Fisher’s exact test were also used to analyze the association between different parameters and recurrence rate with 2 years after hepatectomy, then variables with p < 0.05 were subjected to logistic binary regression. Kaplan-Meier method and log-rank test were used to compare the TTP and overall survival (OS) between different groups. Univariate Cox regression model was used to analyze the association between different parameters and TTP/OS. Variables with p < 0.05 in univariate analysis were subjected to multivariate Cox regression model. The area under the curve (AUC) of receiver operating characteristic curve (ROC) was calculated to evaluate the predictive value for recurrence within 2 years after hepatectomy of the predictive model, TERT mutations and Barcelona clinic liver (BCLC) stage. P < 0.05 was considered significant.
In our study, 81 patients were enrolled, including 66 males and 15 females. The average age was 56 years old. All patients were HBsAg positive and were not infected with other hepatitis virus. The Child-Pugh grades of patients were A (n = 79) and B (n = 2). There were 20 patients with preoperative serum AFP level above 400µg/L before surgery. The median maximum diameter of tumor was 5.5cm (range: 1-19cm) and maximum diameter of tumor in 25 patients were larger than 5cm. There were 20 patients with multiple tumors. The BCLC stage of 10, 53 and 18 patients were 0, A and B respectively. All patients underwent curative hepatectomy, and specimens of 36 patients had microvascular invasion (MVI) in para-carcinoma tissue. The clinicopathological information of enrolled patients were summarized in Table 1.
Factors | Number of patients |
---|---|
Age (< 56/≥56) | 41/40 |
Gender (male/female) | 66/15 |
Hypertension (without/with) | 68/13 |
Diabetes (without/with) | 73/8 |
Family history of cancer (without/with) | 53/28 |
History of alcoholism (without/with) | 56/25 |
Anti-hepatitis virus treatment (no/yes) | 41/40 |
Child-Pugh grade (A/B) | 79/2 |
AFP (< 400/≥400µg/L) | 61/20 |
Tumor number (single/multiple) | 61/20 |
Tumor size (≤ 5cm/>5cm) | 56/25 |
BCLC(0-A/B) | 63/18 |
MVI (without/with) | 45/36 |
Edmondson grade (Ⅰ-Ⅱ/Ⅲ-Ⅳ) | 43/38 |
Liver fibrosis (S1S2/S3S4) | 20/61 |
The most commonly mutated genes of enrolled patients were TP53 (n = 46, 56.8%), TERT (n = 34, 42.0%) and CTNNB1 (n = 20, 24.7%). The alteration type of TERT included promoter mutation (n = 31, 91.2%), gene amplification (n = 1, 2.9%), gene arrangement (n = 1, 2.9%) and substitution (n = 1, 2.9%). Among them, the most common alteration was the C > T mutation at -124 bp from the TERT ATG start site (n = 29, 85.3%) (Table 2). The median TMB value was 3.9 mutations/Mb and the 75% TMB was 8.9 mutations/Mb. TMB values lower than 8.9 mutations/Mb were considered as TMB-low (TMB-L) (n = 61) while TMB values higher than 8.9 mutations/Mb were considered as TMB-high (TMB-H) (n = 20).
Number of patients | Type of genetic alterations | Genetic alterations |
---|---|---|
29 | Promoter mutation | c.-124C > T |
1 | Promoter mutation | c.-146C > T |
1 | Promoter mutation | c.-57A > C |
1 | Substitution | c.425G > A |
1 | Gene amplification | amplification |
1 | Gene arrangement | SLC12A7/TERT |
In this part, we analyzed the correlation between TERT mutations and parameters of HCC patients including twelve clinicopathological characteristics, TP53 mutations, CTNNB1 mutations and TMB. By Chi-square test or Fisher exact test, we didn’t find significantly correlation between TERT mutations and these parameters of patients (p > 0.05) (Table 3).
Factors | TERT | |||
---|---|---|---|---|
WT | Mutant | χ2 | P | |
Age (< 56/≥56) | 22/15 | 19/15 | 0.650 | 0.420 |
Gender (male/female) | 39/8 | 27/7 | 0.166 | 0.683 |
Hypertension (without/with) | 38/9 | 30/4 | 0.789 | 0.372 |
Diabetes (without/with) | 43/4 | 30/4 | N/A | 0.715 |
Family history of cancer (without/with) | 30/17 | 23/11 | 0.127 | 0.721 |
History of alcoholism (without/with) | 33/14 | 23/11 | 0.061 | 0.805 |
AFP (< 400/≥400µg/L) | 36/11 | 25/9 | 0.100 | 0.752 |
Tumor number (single/multiple) | 38/9 | 23/11 | 1.850 | 0.174 |
Tumor size (≤ 5cm/>5cm) | 35/12 | 21/13 | 1.492 | 0.222 |
BCLC(0-A/B) | 39/8 | 24/10 | 1.752 | 0.186 |
MVI (without/with) | 29/18 | 16/18 | 1.713 | 0.191 |
Edmondson grade (Ⅰ-Ⅱ/Ⅲ-Ⅳ) | 27/20 | 16/18 | 0.855 | 0.355 |
TP53 (WT/Mutant) | 24/23 | 11/23 | 2.815 | 0.093 |
CTNNB1 (WT/Mutant) | 37/10 | 24/10 | 0.702 | 0.402 |
TMB (< 8.9/≥8.9 mutations/Mb) | 36/11 | 25/9 | 0.100 | 0.752 |
During the follow-up period, 40 patients had tumor recurrence, and among them 39 patients had tumor recurrence in the first 2 years after hepatectomy. Compared to patients without TERT mutation, the proportion of recurrence within the first 2 years after hepatectomy was higher in patients with a TERT mutation (64.7% vs. 36.2%, p = 0.011). Besides, the recurrence rate within 2 years after hepatectomy was higher in patients with BCLC stage B than others (72.2% vs. 41.3%, p = 0.020). By logistic binary regression, we found that TERT mutations (HR: 2.985, 95%CI: 1.158–7.692, p = 0.024) and BCLC stage B (HR: 3.326, 95%CI: 1.019–10.856, p = 0.046) were independent risk factors for recurrence within 2 years after hepatectomy (Table 4).
Variables | Univariate Analysis | Multivariate Analysis | ||||
---|---|---|---|---|---|---|
n/n | χ2 | P | HR | 95%CI | P | |
Age(< 56/≥56) | 22/17 | 1.010 | 0.315 | |||
Gender (male/female) | 32/7 | 0.016 | 0.899 | |||
Hypertension (no/yes) | 35/4 | 1.873 | 0.171 | |||
Diabetes (no/yes) | 34/5 | N/A | 0.472 | |||
Family history of cancer (no/yes) | 23/16 | 1.387 | 0.239 | |||
History of alcoholism (no/yes) | 28/11 | 0.249 | 0.618 | |||
Anti-hepatitis virus treatment (no/yes) | 20/19 | 0.013 | 0.908 | |||
AFP (< 400/≥400µg/L) | 26/13 | 3.021 | 0.082 | |||
BCLC(0-A/B) | 26/13 | 5.372 | 0.020 | 3.326 | 1.019–10.856 | 0.046 |
Liver fibrosis (S1S2/S3S4) | 7/32 | 1.839 | 0.175 | |||
Edmondson grade (Ⅰ-Ⅱ/Ⅲ-Ⅳ) | 21/18 | 0.017 | 0.895 | |||
MVI (no/yes) | 19/20 | 1.424 | 0.233 | |||
TP53 (WT/Mutant) | 17/22 | 0.004 | 0.947 | |||
TERT (WT/Mutant) | 17/22 | 6.435 | 0.011 | 2.985 | 1.158–7.692 | 0.024 |
CTNNB1 (WT/Mutant) | 27/12 | 1.494 | 0.222 | |||
TMB (< 8.9/≥8.9Muts/Mb) | 28/11 | 0.499 | 0.480 |
Kaplan-Meier analysis and log-rank test were used in survival analysis. In analysis of TTP, the 75% TTP was 11.9 months and 5.4 months in patients with a TERT mutation and others respectively. The difference between two groups was statistically significant (p = 0.003) (Fig. 1). In analysis of OS, the 75% OS of patients with a TERT mutation was 36.9 months, while patients without TERT mutation didn’t reach the 75% OS. There were significant difference between two groups (p = 0.013) (Fig. 2). Sixteen variables were examined in cox regression model to reveal the correlated risk factors for TTP and OS. For TTP, serum AFP level ≥ 400µg/L, BCLC stage B and TERT mutations were significantly correlated with shorter TTP of HCC patients in the univariate analysis (p < 0.05). Multivariate analysis showed that TERT mutations (HR: 2.245, 95%CI: 1.185–4.252, p = 0.013) and BCLC stage B (HR: 2.132, 95%CI: 1.082–4.198, p = 0.029) were independent risk factors for poor TTP (Table 5). For OS, serum AFP level ≥ 400µg/L, MVI, Edmondson grade Ⅲ-Ⅳ, BCLC stage B and TERT mutations were significantly correlated with shorter OS of HCC patients in the univariate analysis. Multivariate analysis showed that BCLC stage B (HR: 4.071, 95%CI: 1.266–13.099, p = 0.019) and MVI (HR: 5.040, 95%CI: 1.090-23.298, p = 0.038) were independent risk factors for OS of HCC patients underwent hepatectomy while the TERT mutation was not (p = 0.063) (Table 6).
Variables | Univariate Analysis | Multivariate Analysis | ||||
---|---|---|---|---|---|---|
HR | 95%CI | P | HR | 95%CI | P | |
Age(< 56/≥56) | 1.626 | 0.868–3.047 | 0.129 | |||
Gender (male/female) | 0.918 | 0.406–2.076 | 0.837 | |||
Hypertension (no/yes) | 0.533 | 0.189–1.498 | 0.232 | |||
Diabetes (no/yes) | 1.448 | 0.566–3.701 | 0.440 | |||
Family history of cancer (no/yes) | 1.023 | 0.572–1.830 | 0.939 | |||
History of alcoholism (no/yes) | 1.373 | 0.729–2.587 | 0.327 | |||
Anti-hepatitis virus treatment (no/yes) | 1.048 | 0.564–1.949 | 0.882 | |||
AFP (< 400/≥400µg/L) | 1.977 | 1.018–3.841 | 0.044 | - | - | 0.199 |
BCLC(0-A/B) | 2.504 | 1.285–4.878 | 0.007 | 2.132 | 1.082–4.198 | 0.029 |
Liver fibrosis (S1S2/S3S4) | 1.363 | 0.628–2.960 | 0.433 | |||
Edmondson grade (Ⅰ-Ⅱ/Ⅲ-Ⅳ) | 1.030 | 0.552–1.921 | 0.926 | |||
MVI (no/yes) | 1.623 | 0.872–3.021 | 0.126 | |||
TP53 (WT/Mutant) | 0.985 | 0.528–1.838 | 0.963 | |||
TERT (WT/Mutant) | 2.502 | 1.335–4.690 | 0.004 | 2.245 | 1.185–4.252 | 0.013 |
CTNNB1 (WT/Mutant) | 1.257 | 0.638–2.476 | 0.509 | |||
TMB (< 8.9/≥8.9Muts/Mb) | 1.369 | 0.683–2.745 | 0.376 |
Variables | Univariate Analysis | Multivariate Analysis | ||||
---|---|---|---|---|---|---|
HR | 95%CI | P | HR | 95%CI | P | |
Age(< 56/≥56) | 0.341 | 0.092–1.258 | 0.106 | |||
Gender (male/female) | 1.473 | 0.397–5.457 | 0.562 | |||
Hypertension (no/yes) | 2.822 | 0.836–9.527 | 0.095 | |||
Diabetes (no/yes) | 1.801 | 0.393–8.262 | 0.449 | |||
Family history of cancer (no/yes) | 1.292 | 0.409–4.079 | 0.662 | |||
History of alcoholism (no/yes) | 0.454 | 0.099–2.072 | 0.308 | |||
Anti-hepatitis virus treatment (no/yes) | 0.700 | 0.222–2.209 | 0.543 | |||
AFP (< 400/≥400µg/L) | 4.755 | 1.508–14.997 | 0.008 | - | - | 0.113 |
BCLC(0-A/B) | 5.967 | 1.891–18.830 | 0.002 | 4.071 | 1.266–13.099 | 0.019 |
Liver fibrosis (S1S2/S3S4) | 0.945 | 0.255–3.499 | 0.933 | |||
Edmondson grade (Ⅰ-Ⅱ/Ⅲ-Ⅳ) | 6.495 | 1.419–29.717 | 0.016 | - | - | 0.162 |
MVI (no/yes) | 6.753 | 1.478–30.854 | 0.014 | 5.040 | 1.090-23.298 | 0.038 |
TP53 (WT/Mutant) | 4.052 | 0.887–18.503 | 0.071 | |||
TERT (WT/Mutant) | 4.532 | 1.225–16.763 | 0.024 | - | - | 0.063 |
CTNNB1 (WT/Mutant) | 1.453 | 0.436–4.836 | 0.543 | |||
TMB (< 8.9/≥8.9Muts/Mb) | 2.224 | 0.705–7.020 | 0.173 |
In order to evaluate the predictive value of 2 independent risk factors (TERT mutations and BCLC stage B) for recurrence within 2 years after curative hepatectomy in hepatitis B-related HCC patients, we established a predictive model. In this model, the study population was divided into 3 groups: group 1 (without any risk factor), group 2 (with 1 risk factor) and group 3 (with 2 risk factors). ROC curves were generated to evaluate the ability of the model to forecast postoperative tumor relapse. As shown in Table 7 and Fig. 3, the model possessed the greater AUC than TERT mutations or BCLC stage (AUC: 0.688 vs. 0.639, 0.688 vs. 0.607, respectively). Therefore, the model had better ability to predict postoperative recurrence than TERT mutations or BCLC stage. Compared to patients in group 1 and group 2, the proportion of recurrence within 2 years after hepatectomy was higher in group 3 (p = 0.006) (Table 8). Besides, Kaplan-Meier curves showed that patients in group 3 had poorer TTP and OS than other groups (p = 0.001, p < 0.001, respectively) (Fig. 4, 5).
Factors | AUC | 95%CI | P |
---|---|---|---|
TERT mutations | 0.639 | 0.517–0.761 | 0.031 |
BCLC stage | 0.607 | 0.483–0.731 | 0.097 |
The model | 0.688 | 0.571–0.804 | 0.004 |
Factor | Group 1 | Group 2 | Group 3 | P |
---|---|---|---|---|
Recurrence within 2 years after surgery | 12 | 19 | 8 | 0.006 |
No recurrence within 2 years after surgery | 27 | 13 | 2 |
TERT has been reported to be the most common mutated gene in HCC patients, with the frequency ranged from 51–60% [11–14]. In this study, the mutation frequency of TERT was 42.0%, which was lower than TP53 (56.8%). The difference between our study and previous studies may be attributed to the different hepatitis backgrounds. In previous studies, HCV infection was the most common in enrolled patients [11–14]. While in our study, all patients were HBsAg positive and without HCV infection. HBV could active the telomerase by inverting into the TERT, which provided an alternative mechanism of reactivating telomerase and might cause the reduced rate of TERT mutations [21, 22].
In our study, TERT mutations consisted of promoter mutation, substitution, amplification and rearrangement. TERT promoter mutation (85.3%) was the most common alteration, which was consistent with results of a previous study [22]. The hot spots of TERT promoter mutation were situated at the − 124 bp and − 146 bp from the TERT ATG start site [22]. The most frequent mutation was c.-124C > T (85.3%) in our cases. But we only found one case with mutation of c.-146C > T. Except for the two hot spots, we also found a A > C substitution at the − 57 bp, which had been reported in melanoma [23]. Other alterations of TERT gene just occupied a small proportion.
Telomeres are DNA-protein structure located at the end of chromosomes. Telemeres shorten accelerates along with the round of cell division and leads to cell death finally [24]. Cancer cells keep continuous division by reactivation of telomerase and maintaining of telomere length. Telomerases are in charge of the telomere synthesis, which consist of TERT, telomerase RNA component and core enzyme [16]. Reactivation of telomerase is a crucial event in hepatocarcinogenesis [16]. There are four mechanisms of telomerase reactivation: (a) thourgh TERT promoter mutations, (b) through HBV insertion into promoter, (c) through TERT amplification and (d) through TERT translocation [15]. TERT controlled the activity of telomerase and is suppressed in most cells of adults [16]. TERT mutations would lead to the overexpression of TERT [25–27].
In this study, we didn’t find the significant correlation between TERT mutations and clinicopathological characteristics of HCC patients such as tumor size, tumor number, MVI, BCLC stage and differentiation. The result was similar with the researches of Kwa et al [28] and Lee et al [29]. Previous studies have reported that TERT mutations could lead to overexpression of TERT in several cancers. For example, transcriptional activity of TERT with c.-124C > T, c.-146C > T or c.-57A > C was higher than wild type in melanoma [23, 30]; Cao et al [31] reported that amplification of TERT increase the expression of TERT; Bayard et al [32] reported that rearrangement of TERT could induce overexpression of TERT. In survival analysis, we found that TERT mutations correlated with shorter TTP and OS of HCC patients underwent hepatectomy. We inferred that this result might be related with the overexpression of TERT caused by TERT mutation. Jeong et al [33] reported that TERT overexpression predicted poor intrahepatic recurrence free survival, this result was similar with ours. Besides, we found that BCLC stage and TERT mutations were independent risk factors for recurrence within 2 years after hepatectomy. A predictive model based on the BCLC stage and TERT mutations showed better predictive value for early recurrence after curative hepatectomy than any single factor. Patients with both BCLC stage B and TERT mutations had poorer prognosis than others. The predictive model may help us to identify patients with high risk of recurrence after hepatectomy. For these patients, more frequent surveillance and intervene in time was of vital importance.
This study has some limitations. First, this is a monocenter and retrospective study, multicenter and prospective studies are needed to acquire more representative and convincing results. Second, all patients enrolled were with HBV infection and without other hepatitis virus. Thus, our result may not applicable to HCC patients infected with other hepatitis virus. Third, due to the limitation of sample size, we couldn’t analyze the correlation between different mutations of TERT and prognosis of HCC patients.
In conclusion, promoter mutation was the most common alteration in TERT. TERT mutations and BCLC stage B were significantly associated with a high risk of early recurrence and predicted poor TTP in HCC patients underwent curative hepatectomy. The predictive model based on the two factors had better ability to predict short postoperative recurrence of HCC patients than TERT mutations or BCLC stage.
Funding
This study did not receive any funding.
Acknowledgments
Thanks to OrigiMed for the technical support with NGS.
Competing interests
The author(s) declare no competing interests.
Data Availability
The datasets analyzed during the current study are available from the corresponding author on reasonable request.
Author contributions
Provision of study material or patients: Liqun Wu; Collection and assembly of data: Liqun Wu, Kangjian Song, Junyu Huo, Fu He, Qingwei Zhu; Data analysis and interpretation: Kangjian Song, Junyu Huo; Manuscript writing: Kangjian Song. All authors contributed to conception and design; took part in drafting and revising the article; gave final approval of the version to be published; and agreed to take responsibility and be accountable for the consents of the article.
Ethics approval
All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards. The study was approved by the Ethics Committee of the Affiliated Hospital of Qingdao University (ethics approval number: QDFYKYLLL-20161212). Informed consent was obtained from all patients included in our study. All participants and contributors of this study have signed informed consent for publication.