Association of renal artery variations with aggressive pathological features and poor prognosis in clear-cell renal cell carcinoma: a computed tomography-based three-dimensional reconstruction study

Introduction

Renal cell carcinoma (RCC) is a malignant tumor arising from renal tubular epithelial cells, accounting for 2–3% of adult malignancies, and representing one of the most common cancers of the urinary system (1). Globally, approximately 403,000 new cases and 175,000 deaths from renal cancer occur annually, with incidence rates continuing to rise (2). Advances in imaging technology have improved the early detection of renal cancer (3). Among its histological subtypes, clear-cell RCC (ccRCC) is the most prevalent, accounting for 70–80% of all renal cancers (4). Compared to papillary and chromophobe RCC, ccRCC is associated with more aggressive biological behavior and a poorer prognosis (5). Early-stage renal cancer is often curable through surgical intervention, while advanced-stage renal cancer generally carries a worse prognosis (6).

As a highly vascularized solid tumor, the initiation and progression of ccRCC rely heavily on adequate blood perfusion, which is primarily supplied by the renal arterial system. An abundant and aberrant tumor vascular network is one of the most distinguishing pathological hallmarks of ccRCC and serves as a critical basis for its biological behavior. At the molecular level, the pathogenesis of ccRCC is typically driven by the inactivation of the von Hippel-Lindau (VHL) tumor suppressor gene (7). Loss of VHL function leads to the aberrant stabilization and accumulation of hypoxia-inducible factors (HIFs) under normoxic conditions, creating a state of “pseudo-hypoxia” (8). The constitutive activation of HIFs further induces the expression of multiple downstream target genes, particularly vascular endothelial growth factor, thereby significantly promoting tumor angiogenesis, metabolic reprogramming, and cell proliferation, driving the initiation, progression, and metastasis of ccRCC (9).

The molecular mechanisms underlying ccRCC angiogenesis have been extensively examined; however, research on the anatomical basis of tumor blood supply—specifically, whether variations in renal artery anatomy affect tumor biological behavior and disease progression—remains limited. Previous studies indicate that the incidence of renal artery variations is higher in patients with renal cancer (17.5–25.1%) than in the general population (14.01%) (10). However, the formation mechanisms of renal artery variations and their potential role in tumorigenesis remain unclear. Some researchers postulate that renal artery variations might be associated with genetic mutations and could provide collateral blood supply to the tumor, potentially influencing tumor growth, pathological features, and prognosis (11,12). Nevertheless, current research on renal artery variations has primarily focused on their impact on surgical approach selection, operative difficulty, and postoperative complications, while their potential role in ccRCC tumor biology has been largely overlooked. Further, most previous studies have focused on vascular anatomical variations in the general population or non-tumor kidneys; however, no systematic analysis has been conducted on the distribution characteristics and clinical significance of renal artery variations in ccRCC patients specifically.

Traditional two-dimensional imaging techniques, such as computed tomography (CT) and magnetic resonance imaging (MRI), provide fundamental anatomical information regarding the kidney and vasculature. However, their limited spatial resolution often hinders the accurate assessment of renal artery variations and their spatial relationship with the tumor (13). In recent years, CT-based three-dimensional (3D) visualization technology has emerged as a powerful tool. By presenting the renal vascular anatomy, variations, and the spatial relationships among the tumor, vasculature, and collecting system in a multi-angular and volumetric manner, this technology significantly enhances the efficacy of preoperative assessment (14). This technology enables surgeons to anticipate technical challenges, identify optimal resection planes, and optimize selective clamping strategies, thereby bolstering the surgeon’s confidence in defining complex relationships between the tumor and critical structures such as the collecting system or segmental vessels (15). A recent large-scale, multi-institutional propensity score-matched analysis demonstrated the tangible clinical benefits of this approach: patients managed with 3D reconstruction models had significantly lower rates of major postoperative complications (3.8% vs. 9.5%, P=0.04), smaller declines in the estimated glomerular filtration rate (−5.6% vs. −10.5%, P=0.002), and higher rates of achieving the Trifecta outcome (55.7% vs. 45.1%, P=0.005) compared to those who received standard care (16).

Based on the hypervascular nature of ccRCC, we hypothesized that renal artery variations—specifically ARAs—may function as collateral vascular supply channels. We postulated that this additional blood supply might support high tumor metabolic demands and be associated with pathological aggressiveness (e.g., a larger tumor size, a higher nuclear grade, and necrosis), and consequently correlate with poorer survival outcomes. In the present study, CT-based 3D visualization technology was used to systematically analyze the characteristics of renal artery variations in patients with ccRCC and to examine their associations with clinicopathological features and survival prognosis. This study aimed to deepen the understanding of the potential clinical significance of renal artery variations in tumor progression, providing valuable data to support the formulation of personalized treatment strategies. We present this article in accordance with the STROBE reporting checklist (available at https://qims.amegroups.com/article/view/10.21037/qims-2025-aw-2447/rc).

Methods

Study population

This retrospective study analyzed clinical data from 386 patients diagnosed with ccRCC and treated at the Department of Urology, The First Hospital of Shanxi Medical University, between January 2019 and December 2024. Clinical and pathological variables included sex, age, body mass index (BMI), tumor location, maximum tumor diameter, tumor growth pattern, World Health Organization/International Society of Urological Pathology (WHO/ISUP) grading, and comorbidities such as hypertension, diabetes mellitus, cerebrovascular disease, and cardiovascular disease. The inclusion criteria were as follows: (I) the presence of a unilateral, solitary tumor; (II) undergoing urological CT or abdominopelvic CT (unenhanced and contrast-enhanced) with 3D reconstruction that clearly delineates the renal artery and its branches; (III) pathologically confirmed ccRCC via renal biopsy or postoperative histopathology; and (IV) availability of complete follow-up data. The exclusion criteria were as follows: (I) incomplete clinical data; (II) the presence of other concurrent malignancies; and/or (III) prior treatment with radiotherapy, chemotherapy, or other anti-tumor immunotherapies.

The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Ethics Committee of the First Hospital of Shanxi Medical University (No. [2021] K048), and the requirement of individual consent for this retrospective analysis was waived.

Imaging and anatomical evaluation

This study analyzed preoperative CT scans from all enrolled patients, using 3D visualization reconstruction to assess the kidney and its vascular structures, as illustrated in Figure 1. Based on the classification by Satyapal et al. (17), an accessory renal artery (ARA) was defined as one or more arteries arising from the abdominal aorta or its branches (e.g., the superior mesenteric artery, inferior mesenteric artery, or common iliac artery) in addition to the main renal artery. A prehilar branching artery (PBA) was defined as one or more branches arising from the renal artery trunk before entering the renal hilum. The classification of renal artery variations was based on the method described by Wu et al. (18), as illustrated in Figure 2: (i) Type I: PBA; (ii) Type II: ARA; (iii) Type III: a combination of Types I and II; and (iv) Type IV: rare variations, such as arterial branches connecting the abdominal aorta and the renal artery trunk, forming an aorta-renal artery bridge. The renal artery trunk was defined as the arterial segment extending from its origin to the first branching point. According to the classification by Weld et al. (19), the anterior renal artery is a branch arising from the renal artery trunk, typically dividing into multiple segmental arteries that supply the renal parenchyma.

Figure 1 Renal artery variations in a patient with ccRCC. CT images (A-C) and 3D visualization models (D) illustrate RAV in a 54-year-old male patient with ccRCC. The tumor measured 3.2 cm in maximum diameter, had a volume of 47.22 cm³, and was classified as WHO/ISUP grade 2. The yellow arrow indicates the PBA 1, the blue arrow indicates the PBA 2, and the green arrow indicates an ARA. 3D, three-dimensional; ARA, accessory renal artery; ccRCC, clear-cell renal cell carcinoma; CT, computed tomography; PBA, prehilar branching artery; RAV, renal artery variation; WHO/ISUP, World Health Organization/International Society of Urological Pathology.

Figure 2 Classification method of renal artery variations. Type I: PBA. Type II: ARA. Type III: a combination of Types I and II. Type IV: rare variations, such as arterial branches connecting the abdominal aorta and the renal artery trunk, forming an aorta-renal artery bridge. ARA, accessory renal artery; PBA, prehilar branching artery.

Imaging assessments were independently conducted by experienced radiologists and urologists who were blinded to the patients’ clinical information and pathological outcomes to prevent assessment bias. In cases of discrepancy, a third expert was consulted to reach a consensus. To quantify the reproducibility of the assessment, inter-observer agreement was evaluated using Cohen’s Kappa statistic. A Kappa value greater than 0.85 indicated excellent agreement (20). The following parameters were recorded based on CT images and 3D reconstruction: (I) the number of renal artery trunks, the type of renal artery variation, the number of anterior and posterior renal arteries, and the number of segmental arteries originating from the anterior renal artery; (II) the number of ARAs, their originating arteries, vertebral level of origin, and termination site; (III) the number and termination sites of PBAs; and (IV) the tumor volume.

Pathological evaluation

All pathological evaluations were conducted by pathologists at our institution. The recorded parameters included the maximum tumor diameter, pathological grade, necrosis, hemorrhage, and cystic degeneration. Tumors were graded according to the WHO/ISUP grading system and classified as low-grade (Grades 1 and 2) or high-grade (Grades 3 and 4) ccRCC (5).

Follow-up

Postoperative follow-up was conducted via outpatient visits or telephone interviews, with assessments scheduled every three months during the first three years, every six months in the fourth and fifth years, and annually thereafter (21). The primary endpoint was overall survival (OS), defined as the time (in months) from surgery to death or the last follow-up (December 20, 2024), while the secondary endpoint was progression-free survival (PFS), defined as the time (in months) from surgery to tumor progression, death from any cause, or the last follow-up (December 20, 2024).

Statistical analysis

Continuous variables were assessed for normality using the Shapiro-Wilk test, and homogeneity of variance was evaluated using Levene’s test. Normally distributed data with homogeneous variance were expressed as the mean ± standard deviation and compared between groups using the independent samples t-test. Data that did not meet the assumptions of normality or homogeneity of variance were presented as the median and interquartile range (IQR, P₂₅–P₇₅) and analyzed using the Mann-Whitney U test. Categorical variables were expressed as frequencies and percentages (%), with group comparisons performed using the Chi-squared test or Fisher’s exact test, as appropriate. Patients with missing clinical data or those lost to follow-up were excluded from the study cohort to ensure data completeness. Survival outcomes, including PFS and OS, were estimated using the Kaplan-Meier method, and differences were compared using the log-rank test. To identify independent prognostic factors, univariate and multivariate Cox proportional hazards regression models were used. Variables demonstrating statistical significance (P<0.05) in the univariate analysis were included in the multivariate model. Results were reported as the hazard ratios (HRs) with the 95% confidence intervals (CIs). Survival analyses and visualization were conducted using R software (version 4.2.1) with the “survival” and “survminer” packages. All other statistical analyses were performed using SPSS version 25.0 (IBM Corp., Armonk, NY, USA). A two-sided P value <0.05 was considered statistically significant.

Results

Clinical and pathological characteristics of patients

The detailed patient selection process is illustrated in Figure S1. Ultimately, a total of 386 patients diagnosed with ccRCC were included in the final analysis. The clinical and pathological characteristics of the cohort are summarized in Table 1 and Table S1. The mean age of the patients was 58.44±11.62 years. Of the patients, 222 were male (57.5%) and 164 were female (42.5%). The mean BMI was 24.82±3.44 kg/m². Among comorbidities, hypertension was present in 175 patients (45.3%), diabetes mellitus in 46 patients (11.9%), cerebrovascular disease in 39 patients (10.1%), and cardiovascular disease in 28 patients (7.3%). Regarding the surgical approach, 235 patients (60.9%) underwent partial nephrectomy, and tumor laterality was evenly distributed. The median maximum tumor diameter was 3.5 cm (IQR: 2.5–5.0 cm), while the median tumor volume was 29.39 cm³ (IQR: 11.27–73.71 cm³). Tumor location was evenly distributed, and 81.6% of the tumors exhibited exophytic growth. High-grade tumors accounted for 24.6% of cases, and intratumoral hemorrhage was observed in 23.6% of cases. Tumor necrosis was present in 16.8% of cases, and cystic degeneration was observed in 14.0% of cases.

Analysis of renal artery variation characteristics

Overall analysis of renal artery variations

The assessment of renal artery variation classification demonstrated high reproducibility between the radiologist and the urologist. The calculated Kappa values for all recorded parameters ranged from 0.90 to 0.98, consistently exceeding the pre-defined threshold of 0.85 (Figure S2). The overall characteristics of renal artery variations are summarized in Table 2. In the entire cohort of 386 patients, only 77 (19.9%) presented with classic single renal artery anatomy bilaterally (no variations on either the ipsilateral or contralateral side). Among the contralateral (healthy) kidneys, 52.3% exhibited no renal artery variation, while the corresponding proportion was significantly lower on the tumor-affected side (33.7%), indicating a significantly higher incidence of renal artery variations in the tumor-affected kidneys (χ² =27.396, P<0.001). In relation to renal artery variations, 79.9% of the patients had a single variant artery on the contralateral side, compared with only 56.3% on the tumor-affected side. The proportion of patients with two and three or more variant arteries on the affected side was 28.1% and 15.6%, respectively (χ² =29.503, P<0.001). On the contralateral side, Type I (58.7%) was the most common variation. Conversely, the affected side was dominated by Type II (43.4%), followed by Type III (25.0%) (χ² =34.151, P<0.001). While no significant difference was observed in the number of renal artery trunks between the two sides, the proportion of patients with only one anterior renal artery was significantly higher on the contralateral side (83.4%) than on the tumor-affected side (72.8%, P<0.001). Conversely, the proportion of patients with more than four anterior renal arteries was significantly higher on the tumor-affected side (27.2%) than on the contralateral side (18.9%, χ² =8.129, P=0.017). No significant difference was observed in the number of posterior renal arteries between the two sides (P=0.185).

Analysis of ARA

The prevalence, number, and distribution patterns of ARAs are summarized in Table 3. In relation to the contralateral (healthy) kidneys, 80.3% of the patients had no ARAs, while this proportion was significantly lower on the tumor-affected side (54.7%), indicating a significantly higher prevalence of ARAs on the affected side (χ² =57.860, P<0.001). Among the patients with ARAs, those with a single ARA accounted for 94.7% on the contralateral side, while this proportion was significantly lower on the tumor-affected side (66.9%). Conversely, the proportion of patients with two or more ARAs was significantly higher on the tumor-affected side (33.1%) than on the contralateral side (5.3%, χ² =22.144, P<0.001). On the contralateral side, 96.2% of ARAs originated from the abdominal aorta, while on the tumor-affected side, only 71.8% had the same origin. The remaining 28.2% of ARAs on the tumor-affected side arose from other blood vessels, demonstrating a significant difference in the origin of ARAs between the two sides (χ² =20.751, P<0.001). No significant differences were observed between the two sides in terms of the vertebral level of origin (P=0.202). Regarding the termination sites, although the overall comparison revealed no statistically significant difference between the two sides (P=0.138), the descriptive analysis revealed a clear predisposition for the lower pole. On the contralateral (healthy) side, the majority of ARAs (53.2%) supplied the lower pole, followed by the upper pole (36.7%), and the middle pole (10.1%). Similarly, on the tumor-affected side, the lower pole remained the most frequent termination site (47.3%), followed by the upper pole (32.8%), and the middle pole (19.8%).

Analysis of PBA

The incidence, number, and distribution patterns of PBAs are summarized in Table S2. A comparison between the contralateral (healthy) and tumor-affected kidneys revealed no significant differences in the prevalence, number, or termination site of PBAs. The proportion of patients without PBAs was 66.8% on the contralateral side and 62.4% on the tumor-affected side (χ² =1.638, P=0.201). Among the patients with PBAs, those with a single PBA accounted for 86.7% on the contralateral side and 89.0% on the tumor-affected side (χ² =0.323, P=0.57). No significant differences were observed between the two sides in the distribution of the PBA termination sites. On the contralateral (healthy) side, 55.4% terminated at the upper pole, 35.1% at the middle pole, and 9.5% at the lower pole. While on the tumor-affected side, 52.4% terminated at the upper pole, 34.1% at the middle pole, and 13.4% at the lower pole (χ² =1.204, P=0.548).

Correlation between ARA and clinicopathological characteristics

Presence of ARAs

Several significant differences were observed between the patients with and without ARAs, as summarized in Table 4. The proportion of exophytic tumors was significantly higher in the ARA group compared to the non-ARA group (86.9% vs. 77.3%, χ² =5.881, P=0.015). While there was no significant difference in tumor location distribution between the two groups (P=0.108), the proportion of left-sided tumors was significantly higher in the ARA group than in the non-ARA group (56.0% vs. 39.8%, χ² =10.062, P=0.002). No significant differences were observed in the incidence of intratumoral hemorrhage (P=0.253) or cystic degeneration (P=0.458) between the two groups. However, the incidence of tumor necrosis was significantly higher in the ARA group than in the non-ARA group (24.6% vs. 10.4%, χ² =13.667, P<0.001). The proportion of high-grade tumors (WHO/ISUP Grade 3–4) was significantly higher in the ARA group than in the non-ARA group (24.0% vs. 15.2%, χ² =4.818, P=0.028). The tumors in the ARA group were significantly larger than those in the non-ARA group: median maximum tumor diameter: 4 vs. 3 cm (P<0.001); and median tumor volume: 57.82 vs. 20 cm³ (P<0.001).

Number of ARAs

Significant differences in tumor growth patterns were observed between patients with a single ARA and those with two or more ARAs, as summarized in Table 5. In the single ARA group, 17.1% of tumors exhibited endophytic growth, whereas this proportion was significantly lower in the multiple ARA group (5.2%) (χ² =4.828, P=0.028). Conversely, the proportion of exophytic tumors was significantly higher in the multiple ARA group (94.8%) than in the single ARA group. No significant differences were observed between the two groups in terms of tumor location or laterality (P>0.05). The two groups also showed no significant differences in the incidence of tumor hemorrhage, cystic degeneration, or pathological grading. The incidence of tumor necrosis was significantly higher in the multiple ARA group compared to the single ARA group (51.7% vs. 11.1%, χ² =34.510, P<0.001). The tumors in the multiple ARA group were significantly larger than those in the single ARA group: median maximum tumor diameter: 5.6 vs. 3.5 cm (P<0.001); and median tumor volume: 131.07 vs. 29.44 cm³ (P<0.001).

Tumor feeding by ARA

Among the patients in whom the ARA directly supplied the tumor, significant differences in tumor growth patterns and pathological characteristics were observed, as summarized in Table 6. The proportion of exophytic tumors was significantly higher in the tumor-feeding ARA group compared to the non-tumor-feeding ARA group (91.6% vs. 81.3%, χ² =4.059, P=0.044). No significant differences were observed between the two groups in terms of tumor location or laterality (P>0.05). The tumor-feeding ARA group had a significantly higher incidence of: intratumoral hemorrhage (32.6% vs. 18.8%, χ² =4.319, P=0.038); tumor necrosis (31.6% vs. 16.3%, χ² =5.506, P=0.019); and cystic degeneration (21.1% vs. 8.8%, χ² =5.038, P=0.025). The proportion of high-grade tumors (WHO/ISUP Grade 3–4) was significantly higher in the tumor-feeding ARA group compared to the non-tumor-feeding ARA group (30.5% vs. 16.3%, χ² =4.853, P=0.028). The tumors were significantly larger in the tumor-feeding ARA group than in the non-tumor-feeding ARA group: median maximum tumor diameter (5.25 vs. 4.0 cm, P=0.035); and median tumor volume (101.17 vs. 51.26 cm³, P=0.019).

Prognostic impact of ARA

A Kaplan-Meier survival analysis was conducted to evaluate the association between the presence of ARAs and patient outcomes. The results demonstrated that patients with ARAs had significantly worse outcomes compared to those without ARAs. Specifically, the presence of ARAs was associated with significantly shorter OS (log-rank P<0.001) and PFS (log-rank P<0.001) (Figure 3).

Figure 3 Impact of ARAs on OS and PFS. A Kaplan-Meier survival analysis was conducted to examine the association between ARA and patient prognosis. (A) OS analysis: patients with ARA variation (blue curve) exhibited significantly lower OS compared to those without ARA variation (red curve) (log-rank test, P<0.001). (B) PFS analysis: ARA variation was significantly correlated with PFS (log-rank test, P<0.001). ARA, accessory renal artery; OS, overall survival; PFS, progression-free survival.

To determine whether ARA is an independent prognostic factor, univariate and multivariate Cox regression analyses were performed. For PFS, the univariate analysis indicated that the presence of ARAs (HR =17.16, P=0.006), along with a high WHO/ISUP grade, tumor necrosis, and a larger tumor diameter, were significant risk factors. In the multivariate analysis, after adjusting for these confounding variables, the presence of ARAs remained a significant independent prognostic factor for disease progression (HR =8.33, 95% CI: 1.03–67.44, P=0.047) (Table 7). In relation to OS, the univariate analysis similarly showed that ARA was associated with an increased risk of mortality (HR =15.35, P=0.01). In the multivariate model, while ARA exhibited a high HR of 6.72, the association did not reach statistical significance (P=0.084), potentially due to the limited number of OS events (Table S3). These findings suggest that the presence of ARAs is closely linked to a poorer prognosis and an increased risk of disease progression, highlighting the potential role of vascular anatomy in characterizing tumor biological behavior.

Discussion

Renal artery variation is closely linked to embryonic development. During embryogenesis, the kidneys, adrenal glands, and gonads receive their blood supply from nine pairs of mesonephric lateral arteries, which originate from the dorsal aorta (22). These arteries are categorized into three groups: cranial arteries (first and second pairs), middle arteries (third to fifth pairs), and caudal arteries (sixth to ninth pairs) (22). The renal arteries develop from a pair of middle arteries. As the metanephros ascends, newly formed superiorly positioned arteries progressively develop, while the lower-positioned arteries regress and eventually disappear, leaving a single renal artery and renal vein. However, the incomplete regression of these arteries may result in renal vascular variations, including the presence of ARAs and PBAs.

The most common types of renal artery variations include ARAs and PBAs. Among these, the ARA type is the most frequent and clinically significant variation, with a reported incidence ranging from 11.3% to 59.5% (23). A study of 820 Turkish patients reported an ARA incidence of 27% and a PBA incidence of 26.7%, with both variations being significantly more common in males than females (24). Similarly, an analysis of 200 Nigerian patients reported an overall renal artery variation incidence of 37%, with ARAs accounting for 23% and PBAs for 4.5%, and a higher prevalence once again observed in male patients (25). A systematic review of African populations reported an overall prevalence of ARAs of 19.7%, with males accounting for 70% of cases, significantly exceeding that in females (26). A study based on an Australian population reported that the incidence of ARAs was 22% at the patient level and 12.12% at kidney level, with no significant difference observed between the left and right kidneys (27). Tao et al. reported a prevalence of renal artery variations of 28.0% in a cohort of 378 Chinese patients (28). Conversely, Ge et al. reported a much higher incidence of 60.81% in a cohort of 298 Chinese patients (29). These disparities in reported prevalence are primarily attributed to ethnic differences, geographical factors, sample sizes, and variations in the definition of ARA (25). In our cohort, the prevalence of renal artery variations and ARAs in the contralateral (healthy) kidneys was consistent with these previously established global ranges.

Reports on renal artery variations in patients with renal tumors remain limited. Our findings demonstrated that the incidence of renal artery variations was significantly higher in tumor-affected kidneys than in contralateral (healthy) kidneys, a trend consistent with previous observations. Lv et al. analyzed 99 ccRCC patients and observed a significantly higher renal artery variation incidence in tumor-affected kidneys compared to contralateral (healthy) kidneys, with the incidence of ARAs being significantly higher in the tumor-affected kidneys than in the contralateral kidneys (30). According to the literature, the reported incidence in renal artery variations in Chinese renal tumor patients ranges from 17.5% to 39.3% (12,31-33). However, the prevalence observed in our cohort was notably higher. This discrepancy may be attributed to the broader definition of renal artery variation used in this study, which included both ARAs and PBAs, whereas most previous studies have focused exclusively on ARAs. Additionally, previous studies often had smaller sample sizes, which might have affected the generalizability of their findings. Overall, these findings suggest that renal artery variations are relatively common in renal tumors, and may be associated with tumor vascularization and disease progression.

According to Graves’ description, the renal parenchyma is divided into five segments: apical, superior, middle, inferior, and posterior, each supplied by branches originating from the renal artery trunk. Notably, there is no collateral arterial supply between these segments, as illustrated in Figure 4 (34). The proportion of patients with multiple anterior renal arteries or with more than four anterior renal artery branches was significantly higher on the tumor-affected side. This suggests that renal tumors may receive blood supply from multiple segmental arteries, raising concerns about whether Graves’ segmentation model accurately represents the vascular anatomy in these cases. Supporting this, Macchi et al. analyzed 15 renal specimens and found that only two cases (13%) conformed to Graves’ segmentation pattern (35). Similarly, Borojeni et al., using renal angiography in 60 patients with T1-stage RCC, observed that in 43% of cases (26 patients), the tumor was supplied by two to four segmental renal arteries (36). Further analysis indicated that when tumor-feeding arteries originated from different segmental arteries, the tumor diameter was generally larger, suggesting a link between vascular complexity and tumor growth. From a surgical perspective, the variability in tumor vascularization has critical implications for partial nephrectomy. Selective segmental artery clamping has been proposed as a strategy to enhance surgical exposure and minimize renal ischemic injury. Shao et al. reported that compared to main renal artery clamping, segmental artery clamping significantly reduced intraoperative ischemic injury and improved postoperative renal function (37).

Figure 4 Graves’ description of renal vascular segmentation. A 3D visualization of a normal kidney is shown, including the anterior view (A), anterior view with segmental arteries (B), posterior view (C), and posterior view with segmental arteries (D). The segmental artery distribution patterns are illustrated from different perspectives: anterior view (E), three-quarter view (F), lateral view (G), and posterior view (H). According to Graves’ classification, the kidney is divided into five vascular segments: the apical segment [1], superior segment [2], middle segment [3], inferior segment [4], and posterior segment [5]. 3D, three-dimensional.

There is considerable variability in the reported incidence of left- vs. right-sided ARAs. Some studies have found that left-sided ARAs are more prevalent (28,29,38,39), while others have reported a higher incidence of right-sided ARAs (33,40). In the present study, ARAs were more frequently identified in the left kidney. From an anatomical perspective, the abdominal aorta serves as the primary origin of the renal arteries. The left side of the aorta offers more available space, which may facilitate the formation and development of ARAs. Conversely, the right side is spatially constrained by the inferior vena cava and liver, which may limit ARA formation. From an embryological perspective, asymmetrical development of the left and right kidneys may result in more complex blood supply on the left side, predisposing it to a higher likelihood of ARA formation.

To further investigate renal artery variations between contralateral (healthy) kidneys and tumor-affected kidneys, a detailed analysis was conducted, focusing on both ARAs and PBAs. The incidence of ARAs was significantly higher in the tumor-affected kidneys than in the contralateral kidneys. The number of ARAs was also greater in the tumor-affected kidneys, with a significantly higher proportion of patients having two or more ARAs. The origin of ARAs was more complex in the tumor-affected kidneys, which had a higher proportion of non-abdominal aortic origins. The incidence of PBAs in our study cohort was consistent with previous anatomical and radiological reports (41). However, the comparative analysis revealed no significant differences between the tumor-affected and contralateral (healthy) kidneys in terms of the prevalence, quantity, or termination patterns of PBAs. Unlike ARAs, which typically arise directly from the aorta due to the persistence of embryonic mesonephric vessels, PBAs originate from the main renal artery trunk. This shared origin with the main renal artery suggests that PBAs may be associated with different hemodynamic or biological profiles compared to ARAs. Consequently, while PBAs represent a notable anatomical variation, their clinical relevance regarding ccRCC aggressiveness appears limited. Nevertheless, the preoperative identification of PBAs using 3D visualization remains clinically crucial for ensuring surgical safety, particularly for guiding precise hilar dissection and preventing intraoperative vascular injury during partial nephrectomy.

The abdominal aorta is the most common origin of ARAs (22). However, in cases of exophytic tumor growth, multiple ARAs tend to develop, and due to tumor compression on surrounding organs, these arteries may establish new vascular connections with other arteries. Upper pole ARAs frequently originate from the celiac trunk, superior adrenal artery, or splenic artery. Lower pole ARAs commonly arise from the mesenteric arteries, common iliac artery, or gonadal artery. Interestingly, in this study, most ARAs terminated in the lower pole of the kidney. Previous studies have also suggested that lower pole ARA variations are more common in Caribbean and British populations (23,27). Lower pole ARAs play a crucial role in supplying blood to the upper ureter. Damage to these vessels, especially during surgical procedures, may increase the risk of ureteral ischemia and necrosis. If an ARA crosses the anterior or posterior aspect of the ureter, it may contribute to ureteropelvic junction obstruction by exerting external compression on the ureter (23).

To further investigate the impact of ARAs on ccRCC, this study analyzed the correlation between ARA presence and tumor clinicopathological features. Tumor differentiation grade is a key indicator of tumor aggressiveness and metastatic potential, while maximum tumor diameter and tumor volume generally reflect tumor burden. Lv et al. found that patients with ARAs on the tumor-affected side exhibited larger tumor diameters, higher nuclear grades, and more frequent exophytic growth (30). Collectively, these findings suggest a strong link between the presence of an ARA and increased tumor proliferation and expansion, which may be attributed to the additional vascular channels available for tumor perfusion. However, the number of ARAs was not found to be significantly associated with WHO/ISUP grading in the present study. We hypothesize that while the presence of an ARA is associated with enhanced tumor vascularization, the number of such vessels may primarily correlate with the physical growth rate and volumetric expansion of the tumor, rather than directly reflecting the degree of cellular differentiation.

Different studies have used varying criteria to define endophytic tumors. In this study, we adopted the definition proposed by Black (42), which classifies a tumor as endophytic only if it is completely encapsulated by normal renal parenchyma. Conversely, exophytic tumors typically refer to isolated tumors located on the renal surface, positioned more than 5 mm from the collecting system or kidney surface. ARAs predominantly supply peripheral renal regions, including the cortex and subcapsular areas. Consequently, when a tumor is situated near the renal surface, the presence of an ARA is frequently associated with an additional blood supply, which may facilitate more rapid tumor expansion (11). Tumor proliferation relies on an adequate blood supply to sustain its increasing demand for oxygen and nutrients. During rapid tumor growth, the tumor center often experiences hypoxia, triggering the activation of HIFs, which in turn stimulates angiogenic pathways, leading to the formation of new blood vessels, which further support tumor expansion and invasion (43,44).

This study found that both the presence and number of ARAs were closely associated with tumor necrosis. This phenomenon may be attributed to the uneven distribution of the additional blood supply provided by ARAs, leading to localized areas of insufficient perfusion, which in turn induces hypoxia-driven necrosis (45). Tumor necrosis is widely recognized as a hallmark of tumor aggressiveness and is frequently associated with a poor clinical prognosis (46). Brinker et al. (47) demonstrated that tumors exhibiting necrosis tend to be characterized by higher staging, higher grading, larger tumor size, and increased extrarenal extension. Notably, when an ARA directly supplied the tumor, an increased incidence of cystic degeneration was observed. Cystic degeneration in ccRCC is a characteristic feature of this tumor subtype, and typically results from hemorrhage and necrosis within glycogen- and lipid-rich tumor masses, leading to the formation of pseudocysts. These cystic tumors are often highly malignant, with pathological specimens frequently exhibiting thick and irregular cystic walls (48).

Our survival analysis provided critical insights into the prognostic value of renal vascular variations. Notably, the multivariate Cox regression analysis identified the presence of ARAs as an independent prognostic factor for PFS, even after adjusting for established risk factors such as WHO/ISUP grade and tumor necrosis. This finding strongly supports our hypothesis that ARAs may function as collateral vascular channels, potentially fueling aggressive tumor biological behavior and driving disease progression. Previous studies have identified WHO/ISUP grading, tumor necrosis, tumor volume, maximum tumor diameter, and growth pattern as independent prognostic factors for ccRCC (49-52). Our study complements these findings by demonstrating that vascular anatomy—specifically the presence of ARAs—adds a distinct and significant dimension to risk stratification. Consequently, these patients may benefit from more intensive postoperative surveillance and closer follow-up strategies.

The present study had several limitations. First, it was a single-center retrospective study conducted in a predominantly Chinese population, which may limit the generalizability of the reported incidence of renal artery variations to other ethnic or geographic groups. Second, the limited follow-up duration prevented a comprehensive assessment of long-term prognosis for some patients. Third, our study used the contralateral (healthy) kidneys as an internal control rather than recruiting an external cohort of healthy individuals without ccRCC. While this self-controlled design minimizes genetic and systemic confounders, the absence of a completely healthy external control group limits our ability to strictly define the baseline prevalence in the general population. Finally, it is important to note that the associations identified in this study are correlational in nature; hence, these findings should be validated in future prospective, multicenter trials to confirm their clinical utility and establish causal relationships.

Conclusions

The prevalence of renal artery variations is higher in ccRCC patients, and the presence of ARAs is significantly associated with exophytic tumor growth, larger tumor volume, higher necrosis rates, and elevated WHO/ISUP grading. The multivariate Cox regression analysis identified the presence of ARAs as an independent prognostic factor for PFS, suggesting that this vascular variation is closely associated with the clinical progression and biological aggressiveness of ccRCC. Future studies should expand the sample size and extend follow-up periods to comprehensively evaluate the impact of renal artery variations on ccRCC progression. Additionally, further research should explore the potential molecular mechanisms underlying renal artery variations in ccRCC progression, providing a more precise theoretical foundation for personalized treatment strategies.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://qims.amegroups.com/article/view/10.21037/qims-2025-aw-2447/rc

Data Sharing Statement: Available at https://qims.amegroups.com/article/view/10.21037/qims-2025-aw-2447/dss

Funding: This research was supported by the Shanxi Provincial Postgraduate Practice and Innovation Project (No. 2025SJ203).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://qims.amegroups.com/article/view/10.21037/qims-2025-aw-2447/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Ethics Committee of The First Hospital of Shanxi Medical University (No. [2021] K048) and individual consent for this retrospective analysis was waived.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Sung H, Ferlay J, Siegel RL, Laversanne M, Soerjomataram I, Jemal A, Bray F. Global Cancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. CA Cancer J Clin 2021;71:209-49. [Crossref] [PubMed]
2. Siegel RL, Miller KD, Wagle NS, Jemal A. Cancer statistics, 2023. CA Cancer J Clin 2023;73:17-48. [Crossref] [PubMed]
3. Capitanio U, Bensalah K, Bex A, Boorjian SA, Bray F, Coleman J, Gore JL, Sun M, Wood C, Russo P. Epidemiology of Renal Cell Carcinoma. Eur Urol 2019;75:74-84. [Crossref] [PubMed]
4. Bukavina L, Bensalah K, Bray F, Carlo M, Challacombe B, Karam JA, Kassouf W, Mitchell T, Montironi R, O'Brien T, Panebianco V, Scelo G, Shuch B, van Poppel H, Blosser CD, Psutka SP. Epidemiology of Renal Cell Carcinoma: 2022 Update. Eur Urol 2022;82:529-42. [Crossref] [PubMed]
5. Moch H, Cubilla AL, Humphrey PA, Reuter VE, Ulbright TM. The 2016 WHO Classification of Tumours of the Urinary System and Male Genital Organs-Part A: Renal, Penile, and Testicular Tumours. Eur Urol 2016;70:93-105. [Crossref] [PubMed]
6. Ljungberg B, Albiges L, Abu-Ghanem Y, Bedke J, Capitanio U, Dabestani S, Fernández-Pello S, Giles RH, Hofmann F, Hora M, Klatte T, Kuusk T, Lam TB, Marconi L, Powles T, Tahbaz R, Volpe A, Bex A. European Association of Urology Guidelines on Renal Cell Carcinoma: The 2022 Update. Eur Urol 2022;82:399-410. [Crossref] [PubMed]
7. Shirole NH, Kaelin WG Jr. von-Hippel Lindau and Hypoxia-Inducible Factor at the Center of Renal Cell Carcinoma Biology. Hematol Oncol Clin North Am 2023;37:809-25. [Crossref] [PubMed]
8. Chen X, Ke Y, Huang X, Chen J, Chen J, Chen Z. Immunotherapy in clear cell renal cell carcinoma: current Status, novel Strategies, and future perspectives. Clin Exp Med 2026;26:116. [Crossref] [PubMed]
9. Comprehensive molecular characterization of clear cell renal cell carcinoma. Nature 2013;499:43-9. [Crossref] [PubMed]
10. Li HB, Han QJ, Li ZJ, Zheng PY, Xu P, Shi ZG, Wang XH. Characters of renal tumor combined with renal vascular abnormality and surgical process experience. Lin Chuang Mi Niao Wai Ke Za Zhi 2018;33:401-3,406.
11. Wang Y, Ren JH. Clinical significance and research progress of accessory renal artery. Zhonghua Nei Ke Za Zhi 2022;61:832-5. [Crossref] [PubMed]
12. Bai S, Wen N, Wu B. The research on renal blood vessel variation in patient with renal cell carcinoma. Xian Dai Zhong Liu Yi Xue 2009;17:1523-6.
13. Jia CY, Chen K, Liu Q, Zhong WB, Xu K, Li BK, Liu CX, Li HL. Application of CT-based renal visual three-dimensional reconstruction model in radical nephrectomy for renal cell carcinoma with renal pedicle vascular variation. Guangdong Yi Xue 2017;38:1379-82.
14. Tang G, Liu H, Wang X, Yao H, Wang D, Sun F, Bao X, Zhou Z, Wang J, Wu J. The role of three-dimensional model in preoperative communication before partial nephrectomy and postoperative management. Asia Pac J Oncol Nurs 2023;10:100222. [Crossref] [PubMed]
15. Isotani S, Kimura T, Ogasa T, Kobayashi T, Hiramatsu I, Ieda T, China T, Shimizu F, Nagata M, Nakagawa Y, Ide H, Horie S. Renal Functional Outcomes in Robot-Assisted Partial Nephrectomy with Minimum Layer Resection Using Virtual Three-Dimensional Image Assistance. J Clin Med 2025;14:7133. [Crossref] [PubMed]
16. Michiels C, Khene ZE, Prudhomme T, Boulenger de Hauteclocque A, Cornelis FH, Percot M, Simeon H, Dupitout L, Bensadoun H, Capon G, Alezra E, Estrade V, Bladou F, Robert G, Ferriere JM, Grenier N, Doumerc N, Bensalah K, Bernhard JC. 3D-Image guided robotic-assisted partial nephrectomy: a multi-institutional propensity score-matched analysis (UroCCR study 51). World J Urol 2023;41:303-13. [Crossref] [PubMed]
17. Satyapal KS, Haffejee AA, Singh B, Ramsaroop L, Robbs JV, Kalideen JM. Additional renal arteries: incidence and morphometry. Surg Radiol Anat 2001;23:33-8. [Crossref] [PubMed]
18. Wu Y. Progress in studies of renal artery variations and its clinical significance. Yi Xue Yan Jiu Sheng Xue Bao 2012;25:1225-8.
19. Weld KJ, Bhayani SB, Belani J, Ames CD, Hruby G, Landman J. Extrarenal vascular anatomy of kidney: assessment of variations and their relevance to partial nephrectomy. Urology 2005;66:985-9. [Crossref] [PubMed]
20. McHugh ML. Interrater reliability: the kappa statistic. Biochem Med (Zagreb) 2012;22:276-82.
21. Zhou H, Lv D, Cui F, Gong Qian, Li J, Wen J, Jia M, Kang Y, Rong Y, Zhang W, Shuang W. Prognostic value of the geriatric nutritional risk index in patients with non-metastatic clear cell renal cell carcinoma: a propensity score matching analysis. Nutr J 2024;23:114. [Crossref] [PubMed]
22. Gulas E, Wysiadecki G, Szymański J, Majos A, Stefańczyk L, Topol M, Polguj M. Morphological and clinical aspects of the occurrence of accessory (multiple) renal arteries. Arch Med Sci 2018;14:442-53. [Crossref] [PubMed]
23. Johnson PB, Cawich SO, Shah SD, Aiken W, McGregor RG, Brown H, Gardner MT. Accessory renal arteries in a Caribbean population: a computed tomography based study. Springerplus 2013;2:443. [Crossref] [PubMed]
24. Gümüş H, Bükte Y, Ozdemir E, Cetinçakmak MG, Tekbaş G, Ekici F, Onder H, Uyar A. Variations of renal artery in 820 patients using 64-detector CT-angiography. Ren Fail 2012;34:286-90. [Crossref] [PubMed]
25. Famurewa OC, Asaleye CM, Ibitoye BO, Ayoola OO, Aderibigbe AS, Badmus TA. Variations of renal vascular anatomy in a nigerian population: A computerized tomography studys. Niger J Clin Pract 2018;21:840-6. [Crossref] [PubMed]
26. Abdu SM, Abdu H, Seid E, Assefa EM. Prevalence of Accessory Renal Arteries in Africa: A Systematic Review and Meta-Analysis Using Anatomical Quality Assurance (AQUA) Checklist. Health Sci Rep 2026;9:e71751. [Crossref] [PubMed]
27. Tardo DT, Briggs C, Ahern G, Pitman A, Sinha S. Anatomical variations of the renal arterial vasculature: An Australian perspective. J Med Imaging Radiat Oncol 2017;61:643-9. [Crossref] [PubMed]
28. Tao XF, Zhu JQ, Wu YW, Tang GY, Shi YZ, Zhang L, Lin Y, Wang ZQ. Dual-energy computed tomography angiography for evaluating the renal vascular variants. Chin Med J (Engl) 2013;126:650-4.
29. Ge GH, Xu YM, Huang W, Zhu YL, Lu DY, Feng Y. Significance of multi-slice spiral CTA angiography in the evaluation of renal artery variations. Xian Dai Mi Niao Wai Ke Za Zhi 2012;17:36-8,98.
30. Lv D, Zhou H, Cui F, Wen J, Shuang W. Characterization of renal artery variation in patients with clear cell renal cell carcinoma and the predictive value of accessory renal artery in pathological grading of renal cell carcinoma: a retrospective and observational study. BMC Cancer 2023;23:274. [Crossref] [PubMed]
31. Huo Y, Su Y, Yang DF, Wang L, Wang GX. Surgical technique of renal vascular variation in the process of posterior laparoscopic radical nephrectomy (with 22 cases handling report). Zhonghua Qiang Jing Wai Ke Za Zhi 2011;4:202-4. (Dian Zi Ban).
32. Wu Y, Cheng XD, Cong J, Zhou WQ. Preoperative application of CT angiography for renal artery in retroperitoneal laparoscopic nephron-sparing surgery. Jiangsu Yi Yao 2012;38:847-8.
33. Zhao J, Wang ZX, Song Y, Hong BF, Zhang X, Fu WJ. Application of three-dimensional angiography in the evaluation of renal artery variation before robotic-assisted laparoscopic partial nephrectomy. Xian Dai Mi Niao Wai Ke Za Zhi 2017;22:921-4.
34. GRAVES FT. The anatomy of the intrarenal arteries and its application to segmental resection of the kidney. Br J Surg 1954;42:132-9. [Crossref] [PubMed]
35. Macchi V, Crestani A, Porzionato A, Sfriso MM, Morra A, Rossanese M, Novara G, De Caro R, Ficarra V. Anatomical study of renal arterial vasculature and its potential impact on partial nephrectomy. BJU Int 2017;120:83-91. [Crossref] [PubMed]
36. Borojeni S, Borojeni A, Panayotopoulos P, Bouvier A, Aubé C, Azzouzi AR, Bigot P. Study of Renal and Kidney Tumor Vascularization Using Data from Preoperative Three-dimensional Arteriography Prior to Partial Nephrectomy. Eur Urol Focus 2020;6:112-21. [Crossref] [PubMed]
37. Shao P, Qin C, Yin C, Meng X, Ju X, Li J, Lv Q, Zhang W, Xu Z. Laparoscopic partial nephrectomy with segmental renal artery clamping: technique and clinical outcomes. Eur Urol 2011;59:849-55. [Crossref] [PubMed]
38. Palmieri BJ, Petroianu A, Silva LC, Andrade LM, Alberti LR. Study of arterial pattern of 200 renal pedicle through angiotomography. Rev Col Bras Cir 2011;38:116-21. [Crossref] [PubMed]
39. Raikos A, Paraskevas GK, Natsis K, Tzikas A, Njau SN. Multiple variations in the branching pattern of the abdominal aorta. Rom J Morphol Embryol 2010;51:585-7.
40. Ugurel MS, Battal B, Bozlar U, Nural MS, Tasar M, Ors F, Saglam M, Karademir I. Anatomical variations of hepatic arterial system, coeliac trunk and renal arteries: an analysis with multidetector CT angiography. Br J Radiol 2010;83:661-7. [Crossref] [PubMed]
41. Jarral SA, Fatima T, Mirza TM, Farooq U. Prevalence of anatomical variant renal arterial diameter, early prehilar branching and multiplicity on contrast enhanced computed tomography scans. Ann King Edw Med Univ 2024;30:163-70.
42. Black P, Filipas D, Fichtner J, Hohenfellner R, Thüroff JW. Nephron sparing surgery for central renal tumors: experience with 33 cases. J Urol 2000;163:737-43.
43. Sekino Y, Teishima J, Liang G, Hinata N. Molecular mechanisms of resistance to tyrosine kinase inhibitor in clear cell renal cell carcinoma. Int J Urol 2022;29:1419-28. [Crossref] [PubMed]
44. Badal SAM, Aiken WD, Chin SN. Molecular Targets and Angiogenesis in Renal Cell Carcinoma, A Multitarget Approach: Mini Review. Curr Drug Targets 2017;18:1204-13. [Crossref] [PubMed]
45. Lam JS, Shvarts O, Said JW, Pantuck AJ, Seligson DB, Aldridge ME, Bui MH, Liu X, Horvath S, Figlin RA, Belldegrun AS. Clinicopathologic and molecular correlations of necrosis in the primary tumor of patients with renal cell carcinoma. Cancer 2005;103:2517-25. [Crossref] [PubMed]
46. Zhang L, Zha Z, Qu W, Zhao H, Yuan J, Feng Y, Wu B. Tumor necrosis as a prognostic variable for the clinical outcome in patients with renal cell carcinoma: a systematic review and meta-analysis. BMC Cancer 2018;18:870. [Crossref] [PubMed]
47. Brinker DA, Amin MB, de Peralta-Venturina M, Reuter V, Chan DY, Epstein JI. Extensively necrotic cystic renal cell carcinoma: a clinicopathologic study with comparison to other cystic and necrotic renal cancers. Am J Surg Pathol 2000;24:988-95. [Crossref] [PubMed]
48. Sun Y, Xu B, Zhang H, Li L, Gao Y, Sun M. Renal clear cell carcinoma undergoing cystic change: A cases report and review of the literature. Int J Surg Case Rep 2024;123:110198. [Crossref] [PubMed]
49. Fukatsu A, Tsuzuki T, Sassa N, Nishikimi T, Kimura T, Majima T, Yoshino Y, Hattori R, Gotoh M. Growth pattern, an important pathologic prognostic parameter for clear cell renal cell carcinoma. Am J Clin Pathol 2013;140:500-5. [Crossref] [PubMed]
50. Schiavina R, Borghesi M, Chessa F, Dababneh H, Bianchi L, Della Mora L, Del Prete C, Longhi B, Rizzi S, Fiorentino M, Martorana G, Brunocilla E. The Prognostic Impact of Tumor Size on Cancer-Specific and Overall Survival Among Patients With Pathologic T3a Renal Cell Carcinoma. Clin Genitourin Cancer 2015;13:e235-41. [Crossref] [PubMed]
51. Kuthi L, Jenei A, Hajdu A, Németh I, Varga Z, Bajory Z, Pajor L, Iványi B. Prognostic Factors for Renal Cell Carcinoma Subtypes Diagnosed According to the 2016 WHO Renal Tumor Classification: a Study Involving 928 Patients. Pathol Oncol Res 2017;23:689-98. [Crossref] [PubMed]
52. Chen SH, Xu LY, Wu YP, Ke ZB, Huang P, Lin F, Li XD, Xue XY, Wei Y, Zheng QS, Xu N. Tumor volume: a new prognostic factor of oncological outcome of localized clear cell renal cell carcinoma. BMC Cancer 2021;21:79. [Crossref] [PubMed]