Investigation of CRP and NLR as potential biomarkers for redefining biological borderline resectable pancreatic cancer

Introduction

The 2017 International Consensus Criteria (ICC) redefined borderline resectable pancreatic ductal adenocarcinoma according to its anatomical, biological, and conditional characteristics (1). Resectable (R) pancreatic cancer with markedly elevated carbohydrate antigen 19-9 (CA19-9) level is classified as biological borderline resectable (BR-B) and is associated with poor prognosis (2). However, consensus regarding the optimal preoperative treatment for BR-B remains lacking.

Although elevated CA19-9 levels clearly indicate the need for preoperative intervention (3), other biomarkers that may define BR-B have not yet been identified. C-reactive protein (CRP), a marker of systemic inflammation, and neutrophil-to-lymphocyte ratio (NLR), an index of inflammation and nutritional status, have been reported to be prognostic factors in pancreatic cancer (4–6). Nevertheless, no study has directly evaluated the combined prognostic significance of CA19-9, CRP, and NLR in resected pancreatic cancer. As such, we focused on CRP and NLR-simple, accessible preoperative biomarkers-and investigated their prognostic relevance alongside CA19-9 to develop an improved definition of BR-B.

Materials and methods

Patient demographics and clinical characteristics

Data from consecutive patients diagnosed with R and BR-B pancreatic cancer, who underwent surgery at the Dokkyo Medical University Saitama Medical Center (Saitama, Japan) between March 2014 and June 2024, were evaluated. A total of 154 patients were included, comprising 87 males and 67 females, with a median age of 74 years (range, 69–78 years). Demographic, operative, and postoperative data were collected prospectively. The timing for measuring laboratory parameters, including biomarkers, such as CA19-9, CRP, and NLR, was defined as pretreatment. For the surgery-first group, this was the preoperative timing; for the neoadjuvant chemotherapy (NAC) group, it was pre-NAC timing. Serum CRP and NLR values were obtained before initiation of treatment. In patients presenting with cholangitis or obstructive jaundice, biomarker measurements were performed only after clinical improvement and resolution of biliary inflammation following appropriate biliary drainage. Patients with active cholangitis or unresolved obstructive jaundice at the time of biomarker assessment were not included in the analysis. Patients with missing CRP or NLR data were excluded from subgroup analyses. Computed tomography was routinely performed preoperatively to evaluate local tumor extent. In accordance with the ICC criteria, BR-B was defined as R-pancreatic cancer with a CA19-9 level >500 U/ml, and patients were categorized into R or BR-B groups. The Lewis antigen status was not routinely assessed in this cohort and was therefore not used as an exclusion criterion. NAC was limited to gemcitabine and S-1 (GS) therapy. Patients who received adjuvant chemotherapy were treated with S-1 for 6 months as the standard regimen. Surgery was only performed in patients without distant metastases. This study was approved by the Institutional Review Board of the Dokkyo Medical University Saitama Medical Center (Approval no. 25044), and written informed consent was obtained from all participants. This study was conducted in accordance with the principles of the Declaration of Helsinki.

Statistical analysis

Categorical variables were compared using the chi-squared or Fisher's exact tests, whereas continuous variables were compared using the Wilcoxon rank-sum test. Overall survival (OS) and disease-free survival (DFS) were estimated using the Kaplan-Meier method, and were compared using the log-rank test. For pairwise comparisons of Kaplan-Meier survival curves, all groups were compared with each other using the log-rank test. Bonferroni correction was applied for multiple comparisons. Statistical significance was defined as P<0.0167 for three-group comparisons and P<0.0083 for four-group comparisons. Patients who were alive at the last follow-up visit were excluded. Multivariate Cox regression analysis was used to identify independent prognostic factors and reported as hazard ratio (HR) and corresponding 95% confidence interval (CI). Variables included in the multivariate Cox regression model were selected a priori based on their clinical relevance and previously reported prognostic factors for pancreatic cancer, including age, surgical procedure, and pathological T stage. Receiver operating characteristic (ROC) curve analysis, based on 1-year OS, was performed to evaluate the prognostic performance of CRP level and NLR. Based on the ROC analysis and clinical relevance, cut-off values of 0.3 mg/dl for CRP and 3.13 for NLR were adopted. The area under the ROC curve (AUC) for CRP was 0.661 (95% CI: 0.495–0.826), with a sensitivity of 55.6% and a specificity of 83.7%. The AUC for NLR was 0.602 (95% CI: 0.457–0.747), with a sensitivity of 56.5% and a specificity of 74.3%. Differences with P<0.05 were considered to be statistically significant. Statistical analyses were performed using SPSS version 29 (IBM Corp., Armonk, NY, USA).

Results

Patient characteristics

Data from 154 patients, comprising 128 with R and 26 with BR-B pancreatic cancer, were included in this study. The median age did not differ significantly between the groups (74 vs. [vs.] 75 years), and the sex distribution was comparable. Pancreaticoduodenectomy (PD) was performed in 104 patients, of whom 50 underwent distal pancreatectomy (DP). No significant differences were observed between the R and BR-B groups in preoperative albumin level, tumor size, T stage, lymph node metastasis, NAC, or adjuvant chemotherapy. Adjuvant chemotherapy was administered to 105 of 154 patients (68.2%), including 87 of 128 patients (68.0%) in the R group and 18 of 26 patients (69.2%) in the BR-B group (P>0.999). Among the 102 patients who received S-1, 88 (86.3%) completed the planned course of treatment, including 72 of 85 patients (84.7%) in the R group and 16 of 17 patients (94.1%) in the BR-B group (P=0.455). In contrast, the BR-B group exhibited markedly higher median CA19-9 levels than the R group (1417.0 vs. 48.3 U/ml; P<0.001) (Table I). The CRP and NLR values stratified according high/low groups also exhibited significant differences across clinical variables, consistent with their biological roles. Due to missing data, CRP and NLR analyses were performed in 141 and 128 patients, respectively (Table II).

Table I. Comparison of clinicopathological characteristics between resectable and biological borderline resectable pancreatic cancer.

Table I.

Comparison of clinicopathological characteristics between resectable and biological borderline resectable pancreatic cancer.

Characteristics	Biological borderline resectable (n=26)	Resectable (n=128)	P-value
Sex, n			0.388
Female	9	58
Male	17	70
Age, years	73.5 (67.5–77.5)	74.5 (69.0–78.8)	0.830
Surgical procedure, n			0.496
PD	16	88
DP	10	40
Preoperative CA19-9, U/ml	1417.0 (785.8–2776.0)	48.3 (12.1–127.6)	<0.001
Preoperative CRP, mg/dl	0.07 (0.06–0.14)	0.11 (0.06–0.28)	0.319
Preoperative albumin, g/dl	3.9 (3.3–4.3)	3.9 (3.6–4.3)	0.276
Preoperative NLR	2.9 (1.6–3.6)	2.4 (1.6–3.4)	0.197
Tumor size, mm	23.0 (18.0–30.0)	20.5 (17.0–30.0)	0.506
T stage, n			0.050
T3/4	20	71
T1/2	6	57
Lymph node metastasis, n			0.121
Positive	4	8
Negative	22	120
Neoadjuvant chemotherapy, n			0.359
Positive	6	44
Negative	20	84
Adjuvant chemotherapy, n			>0.999
Positive	17	81
Negative	9	47

[i] Continuous variables are presented as the median (interquartile range). PD, pancreaticoduodenectomy; DP, distal pancreatectomy; CA19-9, carbohydrate antigen 19-9; CRP, C-reactive protein; NLR, neutrophil-to-lymphocyte ratio.

Table II. Clinicopathological characteristics according to CRP and NLR status.

Table II.

Clinicopathological characteristics according to CRP and NLR status.

	CRP			NLR
Characteristics	High (n=31)	Low (n=110)	P-value	High (n=41)	Low (n=87)	P-value
Sex, n			0.545			0.702
Female	12	50		19	36
Male	19	60		22	51
Age, years	74.0 (70.0–78.0)	75.0 (68.0–78.5)	0.785	76.0 (71.5–79.5)	73.0 (66.0–78.0)	0.002
Surgical procedure, n			0.002			0.687
PD	28	67		29	57
DP	3	43		12	30
Preoperative CA19-9, U/ml	67.1 (2.3–325.3)	69.8 (14.4–314.7)	0.964	75.9 (13.5–349.8)	60.3 (13.4–325.3)	0.256
Preoperative CRP, mg/dl	0.66 (0.47–2.62)	0.06 (0.06–0.13)	<0.001	0.17 (0.06–0.50)	0.08 (0.06–0.18)	0.025
Preoperative albumin, g/dl	3.5 (2.9–3.8)	4.1 (3.7–4.3)	<0.001	3.8 (3.2–4.2)	4.0 (3.6–4.3)	0.256
Preoperative NLR	3.0 (1.9–3.7)	2.4 (1.5–3.2)	0.187	3.8 (3.4–5.3)	1.9 (1.4–2.4)	<0.001
Tumor size, mm	20.0 (18.0–30.0)	21.0 (16.5–31.5)	0.925	25.0 (17.0–30.0)	20.0 (17.0–31.0)	0.102
T stage, n			0.221			0.327
T3/4	21	60		29	53
T1/2	10	50		12	34
Lymph node metastasis, n			0.684			0.710
Positive	1	8		3	5
Negative	30	102		38	82
Neoadjuvant chemotherapy, n			0.668			>0.999
Positive	9	39		16	34
Negative	22	71		25	53
Adjuvant chemotherapy, n			0.086			0.175
Positive	25	70		29	50
Negative	6	40		12	37

[i] Continuous variables are presented as the median (interquartile range). PD, pancreaticoduodenectomy; DP, distal pancreatectomy; CA19-9, carbohydrate antigen 19-9; CRP, C-reactive protein; NLR, neutrophil-to-lymphocyte ratio.

Multivariate Cox regression analysis identified elevated CA19-9, CRP, and NLR as independent predictors of poor prognosis (Table III). More specifically: CA19-9 ≥500 U/ml (HR 4.237, 95% CI: 2.140–8.389 [P<0.001]; CRP ≥0.3 mg/dl (HR 2.595, 95% CI: 1.395–4.829 [P=0.003]; and NLR ≥3.13 (HR 2.250, 95% CI: 1.234–4.102 [P=0.008])

Table III. Multivariate Cox regression analysis of prognostic factors.

Table III.

Multivariate Cox regression analysis of prognostic factors.

Characteristics	Hazard ratio	95% CI	P-value
Sex (male vs. female)	0.741	0.414–1.327	0.313
Age (≥75 vs. <75 years)	1.066	0.799–1.421	0.664
Albumin (<4.0 vs. ≥4.0 g/dl)	0.696	0.509–0.950	0.023
CRP (≥0.3 vs. <0.3 mg/dl)	2.595	1.395–4.829	0.003
CA19-9 (≥500 vs. <500 U/ml)	4.237	2.140–8.389	<0.001
NLR (≥3.13 vs. <3.13)	2.250	1.234–4.102	0.008
Tumor size (≥22.5 vs. <22.5 mm)	0.793	0.586–1.073	0.132
T stage (T3/4 vs. T1/2)	1.625	0.865–3.055	0.132
Lymph node metastasis (positive vs. negative)	1.058	0.319–3.506	0.927

[i] CRP, C-reactive protein; CA19-9, carbohydrate antigen 19-9; NLR, neutrophil-to-lymphocyte ratio.

Albumin <4.0 g/dl was marginally associated with poor survival (P=0.023), although the magnitude of the hazard was lower.

Comparison of survival between R and BR-B cases

The median follow-up was 24.8 months (range, 14.1–48.0 months), and none of the patients were lost to follow-up. Kaplan-Meier analysis revealed that patients with BR-B experienced significantly shorter DFS (P<0.001) and OS (P<0.001) than those with R (Fig. 1A and B). This confirms the poor prognosis associated with BR-B classification based solely on CA19-9 levels.

Figure 1. Comparison of survival between R and BR-B pancreatic cancer. (A) DFS and (B) OS curves comparing R (n=128) and BR-B (n=26) groups. Patients with BR-B disease exhibited significantly worse DFS and OS than those with R disease (both P<0.001). BR-B, biologically borderline resectable; DFS, disease-free survival; OS, overall survival; R, resectable.

Survival stratified according to CRP status

To evaluate the added prognostic utility of CRP, patients were categorized into 3 groups: BR-B with high CRP level (n=3); R with low CRP level (n=89); and all remaining patients (others [n=49]). Clear survival differences were observed among these groups (Fig. 2A and B). The BR-B + high CRP group had a significantly poorer DFS than the others group (P=0.0054) and a significantly poorer OS than the others group (P<0.001), and both DFS and OS were also significantly worse than those of the R + low CRP group (both P<0.001). The R + low CRP group exhibited the best prognosis, with a significantly better DFS and OS than the others group (both P<0.001). This indicates that CRP adds substantial stratification beyond CA19-9 alone.

Figure 2. Survival according to the BR-B/CRP classification. Kaplan-Meier curves of (A) DFS and (B) OS among three groups: BR-B with high CRP (n=3), R with low CRP (n=89) and others (n=49). The BR-B + high CRP group showed significantly poorer DFS and OS than both the others group and the R + low CRP group, whereas the R + low CRP group showed significantly better DFS and OS than both other groups (all P<0.0167 after Bonferroni correction). Only patients with available CRP measurements were included in this analysis (n=141). BR-B, biologically borderline resectable; CRP, C-reactive protein; DFS, disease-free survival; OS, overall survival; R, resectable.

Survival stratified according to NLR status

Similarly, patients were categorized into BR-B with high NLR (n=10), R with low NLR (n=76), and other groups (n=42). Survival analysis revealed the same pattern as that observed in CRP analysis (Fig. 3A and B). The BR-B + high NLR group had a significantly poorer DFS than the others group (P=0.0013) and a significantly poorer OS than the others group (P<0.001), and both DFS and OS were also significantly worse than those of the R + low NLR group (both P<0.001). The R + low NLR group exhibited the best prognosis, with a significantly better DFS than the others group (P=0.0042) and a significantly better OS than the others group (P<0.001). These results further demonstrate that the NLR serves as an important biological marker that refines preoperative risk stratification.

Figure 3. Survival according to the BR-B/NLR classification. Kaplan-Meier curves of (A) DFS and (B) OS comparing the BR-B with high NLR (n=10), R with low NLR (n=76) and others (n=42) groups. The BR-B + high NLR group exhibited significantly worse DFS and OS than both the others group and the R + low NLR group, whereas the R + low NLR group had significantly better DFS and OS than both other groups (all P<0.0167 after Bonferroni correction). Only patients with available NLR measurements were included in this analysis (n=128). BR-B, biologically borderline resectable; DFS, disease-free survival; NLR, neutrophil-to-lymphocyte ratio; OS, overall survival; R, resectable.

Combined biomarker score using CA19-9, CRP, and NLR

To assess cumulative biomarker burden, patients were stratified according to the number of elevated biomarkers among CA19-9 ≥500 U/ml, CRP ≥0.3 mg/dl, and NLR ≥3.13. This stratification resulted in 4 groups: 0 factors (n=55); 1 factor (n=44); 2 factors (n=15); and 3 factors (n=2). Survival curves demonstrated a clear trend toward progressively poorer DFS and OS with increasing biomarker burden (Fig. 4A and B). Significant pairwise group differences included the following: DFS, 0 vs. 1 factor (P=0.0037); 2 vs. 3 factors (P=0.0076); OS, 0 vs. 1 factor (P<0.001); 2 vs. 3 factors [P=0.0587 (borderline)]. Patients with no elevated biomarkers experienced the most favorable outcomes, whereas those with 3 factors had the worst survival, reinforcing the additive prognostic impact of CA19-9, CRP, and NLR.

Figure 4. Survival according to the number of elevated biomarkers (CA19-9, CRP and NLR). Kaplan-Meier curves for (A) DFS and (B) OS in patient groups stratified according to the number of high-risk biomarkers (high CA19-9, high CRP and high NLR): 0 factors (n=55), 1 factor (n=44), 2 factors (n=15) and 3 factors (n=2). The 0 factors group exhibited the most favorable survival, whereas the 3 factors group exhibited the poorest outcomes. Significant differences were observed between the 0 factors group and 1 factor group for both DFS (P=0.0037) and OS (P<0.001), and between the 2 factors group and 3 factors group for DFS (P=0.0076). Only patients with complete CA19-9, CRP and NLR data were included in this analysis (n=116). CA19-9, carbohydrate antigen 19-9; CRP, C-reactive protein; DFS, disease-free survival; NLR, neutrophil-to-lymphocyte ratio; OS, overall survival.

In the favorable prognosis subgroup, defined by the absence of elevated biomarkers, patients who underwent NAC exhibited significantly poorer DFS than those who underwent upfront surgery (P=0.0139) (Fig. 5).

Figure 5. DFS according to treatment strategy in patients without high-risk biomarkers. Kaplan-Meier curves of DFS comparing the NAC group (n=23) and the upfront surgery group (n=32) of patients with 0 high-risk biomarkers. The NAC group exhibited significantly poorer DFS than the upfront surgery group (P=0.0139). DFS, disease-free survival; NAC, neoadjuvant chemotherapy.

Discussion

Postoperative survival after resection of pancreatic cancer has improved with the introduction of multiagent adjuvant chemotherapy. Nevertheless, the outcomes for patients with BR-B pancreatic cancer remain poor, and several reports have emphasized the need for preoperative treatment strategies comparable with those used for anatomically borderline resectable disease (2,7). Therefore, a more refined biological definition of tumor resectability is clinically important.

Historically, BR-B pancreatic cancer was defined using criteria proposed in earlier studies, including CA19-9 level ≥200 U/ml (8–10), cancer-related symptoms lasting >40 days (8), and radiological or cytological suspicion of lymph node metastasis at diagnosis (1). However, the 2017 ICC first established CA19-9 levels >500 U/ml as the primary marker of BR-B pancreatic cancer; this threshold has since been widely adopted (1).

Although CA19-9 plays a central role in estimating tumor aggressiveness (3), it remains unclear whether additional biomarkers would further refine the definition of BR-B pancreatic cancer. As such, we focused on 2 easily measurable inflammatory markers-CRP level and NLR-which emerged as independent predictors of poor prognosis in our multivariate analysis and have been implicated in pancreatic cancer biology in the previous literature (4–6,11). Because these markers can be obtained through routine preoperative laboratory investigations, these markers are highly practical and broadly applicable.

Elevated CRP is not merely a bystander but reflects the complex interplay between the host and the tumor. Advanced tumor progression induces systemic inflammation via the upregulation of the interleukin (IL)-1-IL-6 cytokine network and the release of tumor-derived degradation products, leading to enhanced hepatic CRP production. High CRP levels characterize a proinflammatory tumor microenvironment (TME), which facilitates tumor proliferation, angiogenesis, and the metastatic cascade, while also potentially indicating tumor necrosis and a high tumor burden. Furthermore, CRP may be involved in complement activation and opsonization of tumor cells, thereby linking the innate and adaptive immune responses (12–15).

The prognostic significance of NLR in pancreatic cancer lies in its effect on the balance between tumor-promoting inflammation and anti-tumor immunity. Tumor-driven systemic inflammatory responses play a crucial role in malignant transformation by facilitating increased neutrophil infiltration within the TME, which is a hallmark of cancer progression. These neutrophils can suppress the cytotoxic activity of lymphocytes, whose infiltration is essential for a favorable prognosis but is frequently diminished in the advanced stages of disease. Therefore, an elevated NLR characterizes an immune environment that shifts toward tumor promotion and evasion (16–18).

Previous studies have examined pairwise combinations of CA19-9 levels and CRP or NLR. Nurmi et al (19) evaluated a prognostic score incorporating CA19-9 and CRP and demonstrated significantly worse DFS and disease-specific survival in patients with elevated levels of both markers, and significantly better outcomes when both were low. Similarly, Sakamoto et al (20) stratified patients according to NLR and CA19-9 and found that the group with elevated levels of both markers had the worst prognosis, whereas the double-low group had the best. Asaoka et al (21) also reported that elevated CA19-9 levels and NLR were associated with markedly reduced postoperative survival in pancreatic cancer. The relationship between CRP and NLR has also been described, with Schlick et al (22) reporting that prognosis worsens as the number of elevated markers increases (22,23). A unique feature of the present study was the simultaneous integration of CA19-9, CRP, and NLR, thereby capturing the complementary aspects of tumor burden, systemic inflammation, and host immune status. This three-factor approach enabled a more refined prognostic stratification than models based on a single inflammatory marker and identified a subgroup with particularly favorable outcomes in the absence of all elevated biomarkers.

Building on this evidence, our study explored the utility of combining and scoring 3 biological markers-CA19-9, CRP, and NLR-in the preoperative assessment of pancreatic cancer to refine the concept of BR-B disease. Using this three-factor model, we successfully identified the group with poor prognosis. This suggests that patients with multiple elevated biomarkers may benefit from intensified preoperative management, including more aggressive NAC or structured prehabilitation programs. Our findings suggest a practical framework for the preoperative risk stratification of anatomically resectable pancreatic cancers. Patients without elevated biomarker (CA19-9 <500 U/ml, CRP <0.3 mg/dl, and NLR <3.13) levels experienced favorable outcomes and may be suitable candidates for upfront surgery. In contrast, patients with multiple elevated biomarkers experienced substantially poorer survival and may benefit from more intensive preoperative strategies, including NAC and structured prehabilitation. Therefore, this three-factor model may refine the current definition of BR-B pancreatic cancer by incorporating markers of systemic inflammation and immune status in addition to tumor burden. However, this three-factor model should be regarded as exploratory and hypothesis-generating rather than a validated revised BR-B criterion or definitive treatment algorithm. External validation in larger multicenter cohorts is required before clinical implementation. Furthermore, because the model incorporates CRP and NLR, it may identify biologically aggressive diseases, even in Lewis antigen-negative patients who may not express CA19-9 (24). PD was more frequently performed in the CRP-high group, possibly reflecting differences in tumor location and surgical invasiveness. However, the surgical procedure was not identified as an independent prognostic factor in the multivariate analysis, suggesting that the prognostic significance of CRP and NLR was not solely attributable to differences in the surgical procedures. Interestingly, in the favorable prognosis subgroup (i.e., those with no elevated biomarkers), patients who underwent NAC exhibited poorer DFS outcomes than those who underwent upfront surgery (P=0.014). Although limited by the sample size and retrospective design of this study, this finding suggests that NAC may not be beneficial and may even be detrimental in biologically favorable cases. Therefore, biomarker-guided selection may help to avoid overtreatment. However, this observation should be interpreted with caution and a causal relationship should not be established.

The present study had several limitations, the first of which was its retrospective, single-center design and relatively small sample size, particularly the BR-B subgroup, which may have introduced a selection bias. All patients who underwent NAC were treated with GS and adjuvant chemotherapy consisting of S-1 administered for 6 months. However, treatment strategies have not been fully standardized with regard to the indications and timing of NAC, and whether adjuvant chemotherapy should be administered. In addition, the subgroup with three elevated biomarkers was very small (n=2); therefore, the corresponding findings should be considered exploratory and hypothesis-generating rather than definitive evidence and should be interpreted with caution. CRP and NLR are nonspecific biomarkers that may be influenced by factors unrelated to tumor biology, including the timing of measurement, transient inflammatory conditions, and interlaboratory variability. Although biomarker measurements were standardized to the pretreatment period and obtained only after the resolution of cholangitis or obstructive jaundice, residual confounding factors cannot be excluded. Postoperative complications were not included in the present analysis because this study focused on preoperative prognostic factors. External validation in larger multicenter cohorts is required to confirm the generalizability and clinical utility of the proposed three-factor model.

In conclusion, CRP level and NLR were identified as significant prognostic markers for pancreatic cancer. When combined with CA19-9, these markers enabled a more detailed preoperative stratification and may help identify patients requiring intensive treatment. Conversely, patients with favorable biomarker profiles may be candidates for upfront surgery without NAC. Thus, incorporating CA19-9, CRP, and NLR into the biological evaluation of pancreatic cancer may contribute to a more individualized and biologically informed treatment approach.

Acknowledgements

Not applicable.

Funding

Funding: No funding was received.

Availability of data and materials

The data generated in the present study may be requested from the corresponding author.

Authors' contributions

KK conceived and designed the study. KK collected the data, performed the statistical analyses, interpreted the data and drafted the manuscript. HT contributed to data interpretation, critically revised the manuscript for important intellectual content and supervised the study. TT, MT, SM and HI contributed to data acquisition, data interpretation and critical revision of the manuscript. HY contributed to study supervision, data interpretation and critical revision of the manuscript. KK and HT confirm the authenticity of all the raw data. All authors have read and approved the final version of the manuscript.

Ethics approval and consent to participate

The study was approved by the Ethics Committee of Dokkyo Medical University Saitama Medical Center (approval no. 25044; Koshigaya, Japan). The present study was performed in compliance with the regulations and carried out in accordance with the relevant guidelines. Data collection and analysis followed the ethical guidelines of the World Medical Association Declaration of Helsinki. Written informed consent was obtained from all participants.

Patient consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

Glossary

Abbreviations

Abbreviations:

References