Albumin-bilirubin score: a promising predictor of postoperative distant metastasis in patients with colorectal cancer

ABSTRACT

Aim

The objective of this study was to investigate the prognostic significance of the preoperative serum albumin-bilirubin (ALBI) score in patients with colorectal cancer (CRC).

Methods

Preoperative ALBI data from 1506 CRC patients were categorized into high-value and low-value groups, and the clinicopathological characteristics of these groups were compared. A Cox proportional hazards regression model was employed to investigate the risk factors associated with distant metastasis. The Kaplan-Meier method and log-rank test were utilized to analyze the relationship between ALBI scores and 5-year distant metastasis-free survival (DMFS).

Results

The incidence of distant metastasis in group B (ALBI ≥ −2.4) was significantly higher at 28.81% compared to group A (ALBI < −2.4), which had an incidence of 23.36% (p = 0.018). ALBI score ≥ −2.4 was identified as an independent risk factor for postoperative distant metastasis of CRC (HR = 1.251, 95%CI: 1.021–1.533, p = 0.031). The incidence of 5-year DMFS was significantly lower in the ALBI ≥ −2.4 group compared to that observed in the ALBI < −2.4 group (χ² = 5.97, p = 0.015, log-rank test).

Conclusion

The ALBI score of ≥ −2.4 emerges as an independent risk factor for the occurrence of distant metastases following CRC surgery.

Plain Language Summary

This study analyzed the association between the ALBI score (an index of liver function) and postoperative distant metastasis in 1506 colorectal cancer (CRC) patients. Distant metastasis is where cancer has spread from the original tumor to other organs or lymph nodes. Patients with a high ALBI score (≥-2.4) had a significantly higher incidence of distant metastasis and a substantially lower 5-year distant metastasis-free survival (DMFS) compared to those with a low ALBI score (<-2.4). This is the first report linking the ALBI score to distant metastases, potentially guiding patient management.

1. Introduction

Colorectal cancer (CRC) is a prevalent malignancy of the gastrointestinal tract, ranking third in tumor incidence and second in mortality worldwide according to 2020 data [1]. In recent years, there has been a progressive shift toward an earlier age of onset for CRC, accompanied by a significant increase in the incidence of early-onset CRC (below 50 years old) [2–4]. Moreover, it should be noted that cancers of undefined origin may encompass genuine cases of CRC, potentially leading to an underestimation of its true incidence [5]. The 5-year overall survival (OS) rate among Chinese CRC patients stands at 71.8%, which is comparatively lower than the corresponding figure of 78.2% reported in the United States (data from 2006 to 2019) [6]. Despite significant advancements in diagnostic methods and surgical techniques, postoperative metastasis remains a substantial threat to patient health and is a leading cause of poor OS [7]. Therefore, it is crucial to study the risk factors for postoperative metastasis in CRC and enhance the monitoring of high-risk groups to prolong postoperative survival.

The albumin-bilirubin (ALBI) score is calculated based on two serum indexes, albumin and bilirubin. The levels of serum bilirubin and albumin are intricately linked to hepatic synthesis and metabolism, with serum albumin levels additionally serving as an indicator of the body’s nutritional status. ALBI grading was originally employed to evaluate the degree of liver function impairment in patients with hepatocellular carcinoma (HCC) [8]. The ALBI score surpasses other scoring systems in predicting the prognosis of both HCC and chronic liver disease [9,10]. Research indicates that ALBI grading is crucial for assessing the prognosis of various cancers. A high ALBI score is associated with poorer 5-year survival rates in pancreatic cancer and, when combined with CA19–9, enhances prognostic prediction [11]. In esophageal cancer, a high ALBI score significantly impacts OS and recurrence-free survival [12]. Preoperative high ALBI scores are associated with worse outcomes in gastrectomy patients [13]. Recent studies have also indicated a potential correlation between preoperative ALBI scores and the prognosis of CRC. A propensity score matching analysis showed that a high ALBI score was identified as an independent risk factor for disease-free survival and cancer-specific survival [14]. Another study also showed that elevated ALBI score was an independent risk factor for postoperative complications and OS in patients undergoing radical resection of CRC [15]. However, limited research has been conducted on predicting distant postoperative metastases based on preoperative status. Consequently, this study aimed to analyze the correlation between the preoperative ALBI score and postoperative distant metastasis in CRC patients, thereby providing a foundation for personalized management.

2. Patients & methods

2.1. Study population

A total of 1506 patients diagnosed with CRC underwent surgical procedures at the Affiliated Hospital of Nantong University from January 2013 to December 2017. Inclusion criteria encompassed: (1) receipt of radical surgery; (2) pathological confirmation of CRC; (3) absence of metastatic lesions; (4) no history of liver tumor; and (5) availability of complete preoperative data and clinicopathological features. Patients failing to meet the inclusion criteria were excluded from the analysis.

2.2. Clinical parameters and ALBI score

The serum albumin and bilirubin data were collected within 1 week before surgery, and the ALBI score was calculated using the following formula: (Log₁₀ bilirubin (µmol/L) × 0.66) + (albumin (g/L) × −0.0852) [8]. Additionally, demographic data including age, gender, and pathology of the study participants were collected. The stage was classified based on the eighth edition of the Union for International Cancer Control (UICC) tumor-node-metastasis (TNM) classification [16].

2.3. Follow-up and surveillance

The follow-up assessments were conducted at 3, 6, and 12 months post-surgery, followed by annual evaluations thereafter. Distant metastases were detected through imaging techniques and subsequently confirmed by histological analysis, while the 5-year distant metastasis-free survival (DMFS) data were documented. The time interval between radical surgery and the onset of distant metastasis was defined as the metastasis time. The study was followed up until 31 December 2022.

2.4. Statistical analysis

The X-tile software (Yale University, New Haven, CT, USA) was employed to determine the optimal cutoff value of the ALBI score. Categorical variables were compared using Pearson’s chi-squared test and presented as frequencies and percentages. The Cox proportional hazards regression model was employed to investigate the risk factors associated with distant metastasis. Independent variables that demonstrated statistical significance (p < 0.05) in the univariate Cox regression analysis were included in the multivariate Cox regression analysis. The Kaplan-Meier method and log-rank test were used to analyze the correlation between ALBI scores at different levels and 5-year DMFS. Logistic regression analysis was conducted to investigate the correlation between ALBI level and site of metastasis. Statistical analysis was conducted using the Statistical Package for Social Sciences (SPSS) version 20 (IBM Corp., New York, NY, USA), while graphing analysis was conducted using GraphPad Prism version 7.0 (GraphPad Software Inc., San Diego, CA, USA). p < 0.05 were considered statistically significant.

3. Results

3.1. Baseline information

Out of the 1506 patients, 568 were male and 938 were female, exhibited a median age (Q1, Q3) of 66 (58, 75) years. By December 2022, distant metastases had developed in a total of 384 patients, resulting in a metastasis rate of 25.5%.

3.2. Association between ALBI score and histopathological characteristics as well as distant metastasis

The ALBI scores of 1506 patients ranged from −3.61 to −0.85. Based on the X-Tile analysis results, an optimal cutoff value for the ALBI score was determined as −2.4 (Figure 1). Subsequently, the patients were categorized into two groups: group A (ALBI < −2.4) consisting of 916 individuals (60.8%), and group B (ALBI ≥ −2.4) comprising 590 individuals (39.2%).

Figure 1.

Reveal the optimal cut-off value for ALBI scores through X-Tile analysis.

ALBI: albumin-bilirubin.

The composition ratio of group B exceeded that of group A in the following aspects: females (67.8% vs 58.73%, p < 0.001), age ≥60 years (72.54% vs 58.08%, p < 0.001), tumor size >5 cm (17.63% vs 10.59%, p < 0.001), T_3–4 stage (79.49% vs 72.71%, p = 0.003), TNM III-IV (60.78% vs 43.78%, p < 0.001) and vascular invasion (41.69% vs 36.24%, p = 0.034). Furthermore, the incidence rate of postoperative distant metastasis was higher in group B compared to group A (28.81% vs 23.36%, p = 0.018), as presented in Table 1.

Table 1.

Association between ALBI score and histopathological characteristics as well as distant metastasis [n(%)].

Clinical characteristics	A group (< −2.4，n = 916)	B group (≥ −2.4，n = 590)	χ2	p-value
Gender			12.548	<0.001
Male	378 (41.27)	190 (32.2)
Female	538 (58.73)	400 (67.8)
Age (years)			32.484	<0.001
<60	384 (41.92)	162 (27.46)
≥60	532 (58.08)	428 (72.54)
Tumor size (cm)			15.368	<0.001
<5	819 (89.41)	486 (82.37)
≥5	97 (10.59)	104 (17.63)
Primary site			0.094	0.759
Left colon	708 (77.29)	452 (76.61)
Right colon	208 (22.71)	138 (23.39)
Differentiation			0.869	0.351
Well	817 (89.19)	517 (87.63)
Poor	99 (10.81)	73 (12.37)
T stage			8.896	0.003
T1–2	250 (27.29)	121 (20.51)
T3–4	666 (72.71)	469 (79.49)
N stage			1.970	0.16
N0	546 (59.61)	373 (63.22)
N1–2	370 (40.39)	217 (36.78)
TNM stage			13.319	<0.001
I-II	515 (56.22)	270 (52.94)
III-IV	401 (43.78)	310 (60.78)
Perineural invasion			0.872	0.35
No	725 (79.15)	455 (77.12)
Yes	191 (20.85)	135 (22.88)
Vascular invasion			4.508	0.034
No	584 (63.76)	344 (58.31)
Yes	332 (36.24)	246 (41.69)
Distant metastasis			5.613	0.018
No	702 (76.64)	420 (71.19)
Yes	214 (23.36)	170 (28.81)

ALBI:Albumin-bilirubin; TNM: Tumor node metastasis.

3.3. Cox proportional hazard regression analysis of risk factors for postoperative distant metastasis

In univariate analysis, female (HR = 1.251, 95%CI: 1.012–1.548, p = 0.039), poor differentiation (HR = 1.698, 95%CI: 1.296–2.225, p < 0.001), T_3–4 stage (HR = 2.044, 95%CI: 1.543–2.709, p < 0.001), N_1–2 stage (HR = 2.092, 95%CI: 1.711–2.558, p < 0.001), TNM III-IV (HR = 4.118, 95%CI: 3.295–5.147, p < 0.001), perineural invasion (HR = 1.850, 95%CI: 1.419–2.412, p < 0.001), vascular invasion (HR = 1.772, 95%CI: 1.420–2.211, p < 0.001), and an ALBI score ≥ −2.4 (HR = 1.280, 95%CI: 1.047–1.566, p = 0.016) exhibited significant associations with postoperative distant metastasis.

In multivariate analysis, T_3–4 stage (HR = 1.609, 95%CI: 1.205–2.148, p = 0.001), N_1–2 stage (HR = 1.74, 95%CI: 1.395–2.169, p < 0.001), TNM III-IV (HR = 4.026, 95%CI: 3.184–5.09, p < 0.001), vascular invasion (HR = 1.36, 95%CI: 1.009–1.833, p = 0.044) and an ALBI score ≥-2.4 (HR = 1.251, 95% CI 1.021–1.533, p = 0.031) were identified as independent risk factors for postoperative distant metastases in patients with CRC (Table 2).

Table 2.

Univariate and multivariate cox regression analysis of risk factors for postoperative distant metastasis.

Clinical characteristics	Univariate			Multivariate
	HR	95% CI	p-value	HR	95% CI	p-value
Gender (Female/Male)	1.251	1.012–1.548	0.039	1.217	0.982–1.507	0.073
Age (years) (≥60/<60)	0.970	0.789–1.194	0.776
Tumor size (cm) (≥5/<5)	1.113	0.837–1.481	0.462
Primary site (Right/Left)	0.927	0.727–1.182	0.539
Differentiation (poor/well)	1.698	1.296–2.225	<0.001	1.31	0.993–1.729	0.056
T stage (T3–4/T1–2)	2.044	1.543–2.709	<0.001	1.609	1.205–2.148	0.001
N stage (N1–2/N0)	2.092	1.711–2.558	<0.001	1.74	1.395–2.169	<0.001
TNM stage (III-IV/I-II)	4.118	3.295–5.147	<0.001	4.026	3.184–5.09	<0.001
Perineural invasion (yes/no)	1.850	1.419–2.412	<0.001	1.162	0.894–1.511	0.262
Vascular invasion (yes/no)	1.772	1.420–2.211	<0.001	1.36	1.009–1.833	0.044
ALB (g/L) (<35.7/≥35.7)	1.234	0.999–1.523	0.051
TBI (µmol/L) (≥18.5/<18.5)	1.154	0.885–1.504	0.291
ALBI (≥-2.4/<-2.4)	1.280	1.047–1.566	0.016	1.251	1.021–1.533	0.031

HR: Hazard ratio; CI: Confidence interval; TNM: Tumor node metastasis; ALBI: Albumin-bilirubin; ALB: Albumin; TBI: Bilirubin.

3.4. Correlation between ALBI score and 5-year DMFS

The incidence of 5-year DMFS was significantly lower in the ALBI ≥ −2.4 group compared to that observed in the ALBI < −2.4 group (χ² = 5.97, p = 0.015, log-rank test), as depicted in Figure 2.

Figure 2.

The Kaplan-Meier method and log-rank test were used to analyze the correlation between ALBI score and 5-year distant metastasis-free survival.

ALBI: albumin-bilirubin.

3.5. Association between ALBI score and distant metastasis site

Among the 384 patients with postoperative distant metastasis, single-site metastasis was observed in 127 cases (33.07%), while multiple-site metastasis occurred in 257 cases (66.93%). The liver is the most common site of metastasis, followed by the lungs. Additionally, metastasis can occur in the abdominal cavity, pelvic cavity, bones, and other parts. Figure 3 shows the metastatic site in detail. No significant differences in ALBI scores were found between hepatic and nonhepatic metastasis groups, or between pulmonary and nonpulmonary metastasis groups. However, there were statistically significant differences in ALBI scores between single-site and multi-site metastasis groups, as shown in the Supplementary Table . The logistic regression analysis revealed a significant correlation between the ALBI score and multi-site metastasis (OR = 1.562; 95%CI: 1.011–2.415; p = 0.045). However, no specific association with a particular metastatic site was observed, refer to Table 3 for detailed information.

Figure 3.

Distribution of postoperative metastatic sites in CRC patients.

Table 3.

Univariate logistic regression analysis of ALBI score of CRC patients distant metastasis site.

Varlables	Multisite metastasis			Hepatic metastases			Pulmonary metastasis
	OR	95% CI	p-value	OR	95% CI	p-value	OR	95% CI	p-value
ALBI score (≥-2.4 vs. < −2.4)	1.562	1.011–2.415	0.045	0.949	0.635–1.42	0.801	1.454	0.937–2.258	0.095

Reference category is single site metastasis, nonhepatic metastasis, nonpulmonary metastasis.

ALBI: Albumin-bilirubin; OR: Odds ratio; CI: Confidence interval.

4. Discussion

The occurrence and development of tumors involve complex changes in the body, resulting in changes in certain substances in peripheral blood, which can be used as markers for tumor diagnosis and prognosis. Such as carcinoembryonic antigen (CEA), PD-L1, serum albumin, and bilirubin. CEA functions as a broad-spectrum tumor marker; its elevation is frequently linked to tumor malignancy, heightened risk of recurrence, and metastasis [17]. In instances of mismatch repair deficiency, the expression of PD-L1 in immune cells markedly increases, with the level of PD-L1 expression being crucial for assessing tumor immune evasion and prognosis [18]. Serum albumin is pivotal in sustaining nutritional status and immune function, regulating the activity and functionality of immune cells. It facilitates the differentiation, activation, and proliferation of T cells, B cells, and macrophages [19].

Numerous studies have demonstrated a significant correlation between reduced levels of serum albumin and an unfavorable prognosis in various malignancies, including CRC [20–24]. In addition to its association with liver, gallbladder, and hemolysis-related diseases, serum bilirubin serves multiple functions in the human body [25]. It has been reported that mildly elevated serum bilirubin can exert an antioxidant role in combating diseases associated with oxidative stress, such as cardiovascular disease, and cancer [26–28]. However, excessive elevation of bilirubin may induce pro-oxidative and cytotoxic effects on cells [29]. Studies have indicated a correlation between preoperative elevation of total bilirubin and direct bilirubin levels with poorer OS in CRC patients [30]. Hence, the progression of the disease may be correlated with the extent of bilirubin elevation.

The ALBI score integrates variations in albumin and bilirubin concentrations. It can be observed from the calculation formula that an increase in the ALBI score is associated with a decrease in albumin levels or an increase in bilirubin levels. The presence of an elevated ALBI score typically indicates the occurrence of liver damage. In recent years, numerous studies have confirmed its correlation with the prognosis of digestive tract tumors. For example, in pancreatic cancer, individuals with a high ALBI score exhibit significantly diminished 5-year OS rates compared to those with a low ALBI score _[11]. According to Aoyama et al., the 3-year and 5-year OS and relapse-free survival rates of esophageal cancer patients in the high ALBI group were significantly lower than those in the low ALBI group [12]. Szor et al. demonstrated that patients with gastric cancer in the high ALBI rating group had a poor prognosis after radical resection [13]. Additionally, the study conducted by Kinoshita et al. speculated that patients with an ALBI score of grade 2/3 who underwent surgical resection for non-small cell lung cancer exhibited significantly reduced cancer-specific survival compared to those with an ALBI score of grade 1 [31]. Taken together, these findings indicate a positive correlation between elevated ALBI scores and an unfavorable prognosis in patients with tumors.

Although studies have confirmed that the ALBI score is associated with the prognosis of patients with CRC, these studies have only shown an association between higher ALBI scores in CRC patients and lower OS, lower progression-free survival, and higher rates of postoperative complications [32]. The primary objective of this study was to investigate the correlation between the ALBI score and postoperative distant metastasis in patients with CRC. A novel ALBI cutoff value of −2.4 was established to classify patients into high and low ALBI groups. The results demonstrated a significantly higher incidence of postoperative distant metastases in the high ALBI group compared to the low ALBI group. An independent risk factor for postoperative distant metastases in CRC patients was identified as an ALBI score ≥ −2.4. Possible mechanisms underlying these observations may involve malnutrition induced by hypoalbuminemia, which could potentially reduce anti-cancer immune responses [33]. Elevated bilirubin levels may contribute to pro-oxidation and cytotoxic effects, impairing metabolic function and potentially promoting tumor progression [34]. Bilirubin may potentially impact the composition of intestinal microbiota and subsequently exert an influence on the tumor microenvironment in CRC [35]. Moreover, bilirubin facilitates immune cell suppression and dysfunction during tumor metastasis by modulating the function and expression of macrophages, particularly through upregulating PD-L1 expression and downregulating efferocytosis gene expression, thereby promoting tumor progression [36]. Although further investigation is warranted to elucidate the precise underlying mechanism, our study not only enhances the clinical applicability of the ALBI score but also provides a novel perspective for evaluating metastatic risk in CRC.

After conducting further analysis, we have observed significant disparities in ALBI scores between single and multiple distant metastatic sites. These findings suggest that the ALBI score holds valuable potential for assessing the severity of disease in distant metastasis. However, upon examining the association between the ALBI score and various metastatic sites, no substantial correlation was found, indicating a lack of site-specificity about the ALBI score.

In our study, age was not identified as a significant factor influencing distant metastasis in CRC. However, older patients are at a higher risk of postoperative complications, possibly due to the prevalence of multiple comorbidities such as hypertension and diabetes. Despite this, there remains no clear consensus regarding the direct impact of age on postoperative recurrence, metastasis, and survival in CRC [37]. Additionally, this study also revealed that high T stage, high N stage, high TNM stage, and vascular invasion independently served as risk factors for postoperative distant metastases in patients with CRC. These pathological features reflect the extent of tumor invasion, with a higher stage indicating a poorer prognosis [38]. Previous studies have demonstrated the association between multiple pathological features and prognosis in CRC patients. For example, Liu et al. demonstrated that the N1c subgroup of sigmoid cancer exhibited a poorer prognosis in comparison to the N0 subgroup. Furthermore, the higher T stage and N stage were found to be associated with decreased survival rates in carcinoma of sigmoid [39]. In addition, studies have shown that vascular infiltration is associated with early recurrence and distant metastasis of CRC [40], which aligns with our findings. Therefore, during postoperative follow-up, clinicians should prioritize patients presenting these unfavorable pathological features and high ALBI scores by implementing more proactive management strategies to enhance both metastasis-free survival and overall quality of life for individuals diagnosed with CRC.

Our study also demonstrated that there was no significant association observed when separately evaluating serum albumin and bilirubin for distant metastases in patients with CRC. This suggests that the ALBI score may outperform either albumin or bilirubin alone in prognostic assessment of CRC, as it integrates changes in both biomarkers, enabling a more precise evaluation of the systemic metabolic status of tumor patients.

Our study, based on real-world data, substantiates the correlation between preoperative ALBI scores and postoperative distant metastases in patients diagnosed with CRC. The ALBI score is easy to calculate, incurs no additional costs, and has high repeatability. However, our study has several limitations. Firstly, retrospective studies inherently lack the capacity to fully control for all variables. Secondly, our study did not account for patients’ comorbidities, which may introduce a risk of bias due to potential confounding factors. Thirdly, our study focused on the ALBI score; however, we did not compare it with other scoring systems, such as the metastatic colorectal cancer score (mCCS), albumin-total lymphocyte count-RAS index (ALRI), prognostic score for colorectal cancer (PSCRC). This limits the comprehensiveness of the assessment. To more comprehensively evaluate the risk of metastasis, future efforts should focus on developing nomograms that integrate multiple parameters, including patient age, sex, tumor size, stage, and grade of vascular invasion, to better control for variables and reduce bias.

5. Conclusion

Our research has shown that an ALBI score of ≥ −2.4 is associated with a significant increase in the risk of distant metastasis in CRC, and can be used as a promising predictor. Therefore, it is recommended that clinicians adopt more aggressive treatment approaches, enhance the frequency of follow-up visits, and perform more thorough examinations to ensure that patients receive accurate and effective medical care.

Funding Statement

The work was supported by the  Projects of Nantong Health Commission [MS2022107]; Key medical research projects of Jiangsu Provincial Health Commission [ZD2022008]; National College Students’ innovation and entrepreneurship training program [202210304050Z].

Article highlights

• The optimal cutoff value of ALBI score by X-Tile analysis was −2.4.

• The incidence rate of postoperative distant metastasis was higher in group B (ALBI ≥ −2.4) compared to group A (ALBI < −2.4) (28.81% vs 23.36%, p = 0.018).

• In univariate analysis, an ALBI score ≥ −2.4 (HR = 1.280, 95%CI: 1.047–1.566, p = 0.016) demonstrated a significant association with postoperative distant metastasis in patients with CRC.

• In multivariate analysis, an ALBI score ≥ −2.4 (HR = 1.251, 95% CI: 1.021–1.533, p = 0.031) was identified as an independent risk factor for postoperative distant metastases in CRC patients.

• The incidence of 5-year distant metastasis-free survival (DMFS) was significantly lower in the ALBI ≥ −2.4 group compared to that observed in the ALBI < −2.4 group (χ² = 5.97, p = 0.015, log-rank test).

• Logistic regression analysis revealed a significant correlation between the ALBI score and multi-site metastasis (OR = 1.562; 95%CI: 1.011–2.415; p = 0.045).

• Logistic regression analysis revealed no specific association with any particular metastatic site.

• An ALBI score ≥ −2.4 is a significant risk factor for postoperative distant metastasis in CRC.

Author contribution

Conception and design: Xiuying Shi and Hui Cong; Analysis and interpretation of data: Xiuying Shi and Aijun Shi; Drafting and revision of the manuscript: Xiuying Shi, Aijun Shi and Hui Cong; Final approval: All Authors.

Disclosure statement

The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

Writing disclosure

No writing assistance was utilized in the production of this manuscript.

Ethical conduct of research

The authors state that they have obtained appropriate institutional review board approval ((the Ethics Committee of the Affiliated Hospital of Nantong University, China (Approval no: 2022-K057-01)) and/or have followed the principles outlined in the Declaration of Helsinki for all human or animal experimental investigations. In addition, for investigations involving human subjects, informed consent has been obtained from the participants involved.

Financial disclosure

This study was mainly supported by National College Students’ innovation and entrepreneurship training program (202210304050Z), Projects of Nantong Health Commission (MS2022107), Key medical research projects of Jiangsu Provincial Health Commission (ZD2022008). The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/17520363.2025.2455928