Analysis of the positional relationship between the left brachiocephalic vein and its surrounding vessels via computed tomography scan: A retrospective study

Abstract

Objectives

This study aimed to clarify the positional relationship between the left brachiocephalic vein and its surrounding vessels and to analyse the association between this positional relationship and ageing.

Method

Chest contrast-enhanced computed tomography was performed for 100 adults. The contact number between left brachiocephalic vein and surrounding vessels (aorta, brachiocephalic artery, left common carotid artery and left subclavian artery) was determined. The correlations of ageing with the cross-sectional areas of left brachiocephalic vein crossing brachiocephalic artery and left common carotid artery and peripheral end of left brachiocephalic vein were analysed.

Results

LBV was in contact with aorta in 19, brachiocephalic artery in 97, left common carotid artery in 90 and left subclavian artery in 21 patients. There were significant negative correlations of ageing with the cross-sectional areas of left brachiocephalic vein crossing brachiocephalic artery and left common carotid artery and peripheral end of left brachiocephalic vein.

Conclusions

Brachiocephalic artery and left common carotid artery have easy contact with left brachiocephalic vein. There was a negative relationship between the cross-sectional area of left brachiocephalic vein and age.

Keywords

Brachiocephalic vein stenosis occlusion ageing

Introduction

Clinical problems are observed at sites with veins that are compressed by arteries. The left renal vein is located anterior to the abdominal aorta and posterior to the superior mesenteric artery and is often strangulated at this position, possibly leading to nutcracker syndrome, characterised by haematuria and back pain.^1,2 The left common iliac vein is located anterior to the vertebral body and posterior to the right common iliac artery and is easily strangulated; this anatomical relationship may be a cause of deep vein thrombosis.^3,4

The three veins in the upper mediastinum (the left brachiocephalic vein (LBV), right brachiocephalic vein and superior vena cava) are collectively called the central veins and are frequently used as access routes for catheterisation and central venous nutrition.⁵ Central venous stenosis occurs in 25–40% of patients who undergo maintenance dialysis.^6–8 Among the central veins, LBV is a popular choice for implantable access port insertion and pacemaker placement. However, complication rates during such procedures are higher with LBV than with the right brachiocephalic vein.⁹ Similar to the left renal vein and left common iliac vein, LBV is susceptible to compression from its surrounding structures; this may be the cause of complications.

It is known that the aortic arch and its branches (the brachiocephalic artery (BA), left common carotid artery (LCCA) and left subclavian artery (LSA)) are located in a narrow area between the sternum and thoracic vertebral body and are located close to LBV.⁵ This structure may be responsible for the high probability of complications after treatment via LBV. However, no anatomical basis to support this notion has been identified.

Therefore, this study focused on the structural arrangement where LBV is located anterior to the aortic arch and its branches (BA, LCCA and LSA) and posterior to the sternum. Using thoracic contrast-enhanced computed tomography (CT), we analysed the positional relationship between LBV and these structures and revealed the relationship between these structures and age-related changes.

Methods

Participants

We examined 100 participants (mean age: 61.7 ± 17.5 years; 42 women, 58 men) using chest contrast-enhanced CT at the Aichi Medical University Hospital from 2016 January to 2017 February. Regular CT venous phase images were used. We excluded the following patients: those with cancer, thoracic trauma, placement of artefacts in the chest (cardiac pacemaker, central venous catheter and endotracheal tube), mediastinal tumours, pulmonary embolism, acute heart failure, kyphosis and scoliosis; those who had undergone thoracic surgery, mammary tumour resection and orthopaedic surgery; those on dialysis; those with unstable vital signs, those undergoing emergency examination and those for whom dehydration was recorded in their medical history and observed on physical examinations and blood tests.

Analysis of CT images

Two physicians (one radiologist and one vascular surgeon) assessed all patients. All patients underwent a multi-detector row CT scan examination using a SOMATOM Definition AS + CT scanner (Siemens, Healthcare GmbH, Forchheim, Germany) in the supine position. Scanning parameters were as follows: (1) axial images with slice thickness of 1–2 mm; (2) rotation speed of 0.3–0.5 s/rotation and (3) beam pitch of 0.5–1.4. Images were acquired during maximum inspiration and breath-holding. Non-ionic iodinated contrast medium (2.0 mL/kg body weight; Iopamidol, 300–370 mg/mL) was administered through a power injector into the right arm vein at a speed of 2–4 mL/s. The contact number between LBV and surrounding vessels (aorta (Ao), BA, LCCA and LSA) was determined using the axial and sagittal sections of CT images. In the sagittal section where LBV was in contact with each artery, it was confirmed which bone was present in front of LBV. The relationship between the number of LBV contacts with each artery and age was analysed.

The cross-sectional areas of LBV crossing BA and LCCA and at the cross section of the peripheral end of LBV (PLBV) were measured. PLBV was defined on CT image as the region immediately after the point where the subclavian vein and internal jugular vein (IJV) merged. Moreover, BA, LCCA and LSA were examined for any branching abnormalities at their sites of branching off.

Classification of contact patterns between LBV and surrounding vessels

Theoretically, there are 16 patterns for LBV contact with Ao, BA, LCCA and LSA (Table 1). The pattern in which LBV contacts one artery was type I. The pattern in which LBV contacts two arteries was type II. The pattern in which LBV contacts three arteries was type III. The pattern in which LBV contacts all four arteries was type IV. Type V was used when LBV did not contact with any of the four arteries.

Table 1.

Pattern of the left brachiocephalic vein in contact with arteries.

Type	Contacting arteries
Type I
a	Ao
b	BA
c	LCCA
d	LSA
Type II
a	Ao + BA
b	BA + LCCA
c	LCCA + LSA
d	Ao + LSA
e	BA + LSA
f	Ao + LCCA
Type III
a	Ao + BA + LCCA
b	BA + LCCA + LSA
c	Ao + LCCA + LSA
d	Ao + BA + LSA
Type IV	Ao + BA + LCCA + LSA
Type V	None

Ao: aorta; BA: brachiocephalic artery; LCCA: left common carotid artery; LSA: left subclavian artery.

Relationship of age with cross-sectional area and stenosis ratio of the blood vessel

The relationship of age with cross-sectional areas of LBV crossing BA and LCCA and PLBV was analysed. Furthermore, each cross-sectional area was compared. In general, veins become thicker towards the centre. Therefore, the thinnest site in LBV is considered to be PLBV, which is most peripherally located. There is no structure around PLBV that might compress it, and the possibility of stenosis is low. Therefore, it may be suitable to use the cross-sectional area of PLBV as a reference for the cross-sectional area of LBV. The area ratio at each site was measured by dividing the cross-sectional area at the site where LBV crossed BA or LCCA by the cross-sectional area of PLBV. Generally, a stenosis of ≥50% is defined as a significant stenosis in the coronary artery.¹⁰ Therefore, in the present study, a stenosis of ≥50% was considered to indicate an area ratio of ≤0.5. The relationship between age and area ratio was analysed.

Statistical analysis

Data were analysed using Pearson’s correlation coefficient and student’s t-test. GraphPad Prism version 8.2.1 (GraphPad Software, Inc., San Diego, CA) was used for all analyses. A nominal two-sided p < 0.05 was considered significant.

Results

Contact patterns of LBV and the surrounding vessels

The three-dimensional (3D) and axial and sagittal CT images of LBV of a 47-year-old man are shown in Figure 1. Several vessels can be seen adjacent to LBV. In particular, the aortic arch, its branches (BA, LCCA and LSA) and LBV are located in a narrow area between the sternum and thoracic vertebral body. LBV was in contact with the Ao in 19 (19.0%), BA in 97 (97.0%), LCCA in 90 (90.0%) and LSA in 21 (21.0%) patients (including duplication; Figure 2). Theoretically, there are 16 patterns for LBV contact with Ao, BA, LCCA and LSA; however, only seven patterns were observed in this study (Figure 3). In total, 57 patients presented with type II-b. LBV, which is in contact with Ao, was sandwiched between sternum and Ao in 19 of 19 (100%) patients. LBV, which is in contact with BA, was sandwiched between sternum and BA in 97 of 97 (97.0%) patients. Figure 4(a) shows a typical example. LBV, which is in contact with LCCA, was sandwiched between sternum and LCCA in 85 of 90 (94.4%) patients. LBV, which is in contact with LSA, was sandwiched between sternum and LSA in 11 of 21 (52.4%) patients. However, in 5 of 90 (5.6%) patients in whom LBV was in contact with LCCA and 10 of 21 (47.6%) patients in whom LBV was in contact with LSA, the left clavicular sternal end was present anteriorly. The age and rate at which LBV was in contact with arteries were not correlated to each artery (Ao, BA, LCCA and LSA; p > 0.05).

Figure 1.

Three-dimensional (3D) and computed tomography (CT) images of the left brachiocephalic vein (LBV) of a 47-year-old man: (a) 3D, (b–d) axial and (e–g) sagittal CT images of LBV and surrounding arteries.

Figure 2.

Number at which the four arteries contacted with the left brachiocephalic vein.

Figure 3.

Result of pattern of the left brachiocephalic vein in contact with arteries. 1: left brachiocephalic vein; 2: brachiocephalic artery; 3: left common carotid artery; 4: left subclavian artery; 5: aorta.

Figure 4.

Computed tomography (CT) image of the compressed left brachiocephalic vein (LBV) in the axial section: (a) LBV was compressed by the brachiocephalic artery (BA) in a 66-year-old woman. The cross-sectional area at the site where LBV was in contact with BA (arrowhead) was 21.8% of the peripheral end of LBV (PLBV). (b) LBV was compressed by the left common carotid artery (LCCA) in a 47-year-old woman. The cross-sectional area at the site where LBV contacted with LCCA (arrowhead) was 49.8% of PLBV.

Relationship of age with cross-sectional area and stenosis ratio of the blood vessel

There were negative correlations of ageing with the cross-sectional areas of LBV crossing BA and LCCA and PLBV (BA: r = −0.29, p < 0.01; LCCA: r = −0.52, p < 0.001; peripheral end: r = −0.51, p < 0.001; Figure 5). The cross-sectional area of LBV crossing BA and LCCA was not significantly different from that of PLBV (p > 0.05). No correlation was observed between ageing and the ratio of the cross-sectional area of LBV crossing BA and LCCA to the cross-sectional area of PLBV (BA: r = 0.12, p > 0.05; LCCA crossover: r = −0.11, p > 0.05). The cross-sectional area of LBV crossing BA and LCCA was smaller than that of PLBV in 40 and 51 patients, respectively (Figure 6). In total, five and four patients with LBV crossing BA and LCCA, respectively, had ratios ≤ 0.5 (Figure 4). Bifurcation abnormalities of the aortic arch were observed in 10 (10.0%) patients with a common duct at the origin of BA and LCCA (bovine arch) and in 2 (2.0%) patients with a common duct originating at LCCA and LSA. In 12 patients with a bifurcation abnormality in the aortic arch, LBV was in contact with Ao in 1 (8.3%), BA in 12 (100%), LCCA in 11 (91.7%) and LSA in 1 (8.3%) patient (including duplication).

Figure 5.

Relationships of age with the cross-sectional areas of the left brachiocephalic vein (LBV) crossing the brachiocephalic artery (BA) (a) and left common carotid artery (LCCA) (b) and the peripheral end of LBV (c). BA: n = 100, r = −029, p < 0.01; LCCA: n = 100, r = −0.52, p < 0.001; peripheral end: n = 100, r = −0.51, p < 0.001.

Figure 6.

Ratio of the cross-sectional area of the left brachiocephalic vein (LBV) crossing the brachiocephalic artery (BA) and left common carotid artery (LCCA) to that of the peripheral end of LBV (PLBV). The area ratio at each site was measured by dividing the cross-sectional area at the site where LBV crossed BA or LCCA by the cross-sectional area of PLBV.The grey zone shows cases with ratios ≤ 0.5, and there were five and four patients with LBV crossing BA and LCCA, respectively.

Discussion

The present study assessed the positional relationship between LBV and its surrounding vessels as well as the correlation between this positional relationship and age-related changes using chest contrast-enhanced CT. LBV was found in most cases in contact with BA and LCCA, but only in a few cases with Ao and LSA (Figure 2). The cross-sectional area of LBV decreased with age (Figure 5). To the best of our knowledge, this is the first study to evaluate the positional relationship between the aortic arch branches and LBV in non-dialysis adult patients. These anatomical features are possibly involved in the occurrence of complications when LBV is used for central venous access.

In this study, the pattern of the arteries in contact with LBV was highest in 57 patients with type II-b (BA and LCCA) but was less in patients in contact with Ao (Figure 3). Five and four patients had LBV crossing BA and LCCA, respectively, with stenosis of ≥50% (Figure 6). Circulation disorder due to central venous stenosis is one of the complications of inner shunts for blood dialysis. Many studies have focused on LBV compression, stenosis or obstruction in haemodialysis patients.^11–13 Dialysis access-related central venous stenosis occurs in 25–40% of dialysis patients, and stenosis or obstruction in LBV may be caused by venous valves or compression of the sternum and aortic arch.^7,8 Shi et al.¹² have reported that symptomatic venous stenosis or obstruction in dialysis patients with inner shunt in the left upper limb is prevalent in LBV. This included patients with aortic aneurysm in the aortic arch, and some cases of LBV stenosis due to compression from the aortic arch were observed. Therefore, aortic aneurysms may affect LBV stenosis. However, in our study without aortic aneurysm, there were only few cases of contact with Ao and many cases of contact with BA and LCCA.

In studies targeting non-dialysis patients, Oginosawa et al.¹⁴ reported complete venous occlusion associated with the development of collateral circulation in 12 (4.4%) patients and venous occlusion at the site of LBV in 9 (3.3%) patients; these patients were asymptomatic. Guo et al.⁸ reported that 75 (35.4%) of 212 patients with LBV stenosis of >25% were asymptomatic. In the present study, five and four patients had LBV crossing BA and LCCA, respectively, with stenosis of ≥50%, which is considered significant for the coronary arteries; however, all patients were asymptomatic (Figure 4). Significant stenosis of the veins was not defined; therefore, we used a criterion of ≥50% stenosis, which is significant in the coronary arteries, but we could not identify a significant stenosis in the veins. Therefore, numerous non-dialysis patients may be asymptomatic, even if there is obstruction or stenosis in LBV. On the other hand, in our study, age-related changes were not observed in the ratio of the cross-sectional area of LBV crossing BA and LCCA and that of PLBV. Thus, even if a patient is asymptomatic, regardless of age, evaluation of occlusion and stenosis ratio where LBV crosses BA and LCCA before a dialysis shunt is created may contribute to the diagnosis and prediction of the development of central venous stenosis after the introduction of dialysis. Furthermore, the presence of aortic aneurysm may contribute to the diagnosis and prediction of the development of central venous stenosis after the introduction of dialysis.

In several studies, central venous stenosis develops after the placement of the totally implantable venous access port (TIVP) or peripherally inserted central venous catheter (PICC). Song et al.¹⁵ reported the risk factors for central venous stenosis after placing TIVP in patients with breast cancer. They evaluated 191 women with breast cancer; >25% central venous stenosis occurred in 1 (1.1%) of 89 right-sided and in 14 (13.7%) of 102 left-sided patients, and it was significantly more likely to occur in the left side than in the right side. Furthermore, Smith et al.¹⁶ reported that PICCs inserted via the upper limbs have a higher probability of occlusion when inserted from the left side than from the right side. In addition, Paquet et al.¹⁷ reported a higher rate of complications with a PICC inserted from the left upper limb than that inserted from the right upper arm. The cross-sectional area of LBV crossing BA and LCCA, which is in the middle of LBV, was smaller in 40 and 51 patients, respectively, than that of PLBV, where the cross-sectional area is considered the narrowest (Figure 6). This may be correlated with the occurrence of central venous stenosis caused by a TIVP placed via the left IJV or obstruction due to a PICC inserted from the left upper limb.

The cross-sectional area of PLBV and that of LBV crossing BA and LCCA significantly decreased with age (Figure 5). There has been an increase in the number of stents being placed for iliac vein compression syndrome,¹⁸ and reports of stent placement for LBV stenosis or occlusion are also occasionally observed.^6,19,20 Sobrinho and Aguiar²⁰ have reported on stenting for symptomatic superior vena cava syndrome associated with lung malignancy. Stent placement failure was observed in three (25%) of 12 patients in whom the superior vena cava was occluded and in five (83.3%) of six patients in whom the bilateral brachiocephalic veins were occluded. Although the occurrence of brachiocephalic occlusion is rare, the treatment success rates are low. In addition, the treatment success for stent placement in the central vein of dialysis patients is not good compared with that for balloon angioplasty.^21–23 Reduction in the cross-sectional area of LBV with age may have an impact on stent placement failure or restenosis after placement. In the present study, the cross-sectional area of LBV crossing BA was narrower than that of PLBV in 40 patients and the cross-sectional area of LBV crossing LCCA was narrower than that of PLBV in 51 patients (Figure 6). Moreover, there were negative correlations of ageing with the cross-sectional areas of LBV crossing BA and LCCA and PLBV. This result should be taken into consideration when choosing the size for stent placement. Adequate knowledge regarding LBV is essential among radiologists (image diagnosis, stent placement and ballooning), anaesthesiologists and oncologists (central venous catheter placement, TIVP or PICC), cardiologists (pacemaker placement), nephrologists (creation of an arteriovenous shunt for haemodialysis) and vascular surgeons.

The present study has certain limitations. First, only regular venous phase CT, not phlebography or intravenous ultrasonography (IVUS), was performed. Currently, IVUS combined with conventional venography is the gold standard for evaluating venous structures. Use of phlebography and IVUS may have provided a more accurate diagnosis of stenosis in LBV. In addition, only a small number of patients were included in this study, which may have led to an under- or overestimation of the positional relationship between LBV and its surrounding structures. In this study, we measured the cross-sectional area of LBV in the supine position, but the change in body position could have caused changes in the pressure by the surrounding organs as a result of gravity, which may have changed the cross-sectional area of LBV. Furthermore, we evaluated the relationship between LBV and the surrounding arteries but did not consider muscles and nerves. Moreover, the definition of venous stenosis has not been finalised and the accuracy of this measurement method is unverified. Thus, further studies focusing on anatomical structure should be performed.

This study is the first to evaluate the positional relationship between the aortic arch branches and LBV in non-dialysis adult patients. Theoretically, there are 16 patterns for LBV contact with Ao, BA, LCCA and LSA. However, only seven patterns were observed in this study. In most cases, LBV was in contact with BA and LCCA; however, in few cases, it was in contact with Ao and LSA. Moreover, a negative correlation was observed between the cross-sectional area of LBV and age.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Ethical approval

The ethics committee of Aichi Medical University approved this study (permit number: 2016-310).

Guarantor

MN.

Contributorship

HM, YO, MN, TN and HI designed the study. HM, TA, YO, MH and MN performed data analysis. HM and YO wrote the manuscript. All authors reviewed and edited the manuscript and approved the final version of the manuscript.

Acknowledgements

We would like to mention our special thanks to Ikuo Sugimoto, Tetsuya Yamada, Yuki Orimoto, Yuki Maruyama, Yusuke Imaeda and Takashi Ohta for their supportive work.

ORCID iD

Hiroki Mitsuoka

References

Shokeir

el-Diasty

Ghoneim

MA.

The nutcracker syndrome: new methods of diagnosis and treatment. Br J Urol 1994; 74: 139–143.

Uzunlulu

Guzelbey

, et al. The nutcracker syndrome: the usefulness of different MRI sequences for diagnosis and follow-up. Clin Imaging 2019; 55: 144–147.

Sang

, et al. Diagnosis and endovascular treatment of iliac venous compression syndrome. Phlebology 2019; 34: 40–51.

Mitsuoka

Ohta

Hayashi

, et al. Histological study on the left common iliac vein spur. Ann Vasc Dis 2014; 7: 261–265.

Woodhouse

Waheed

Bordoni

Anatomy, thorax, brachiocephalic (innominate) veins. In: StatPearls [Internet]. Treasure Island, FL: StatPearls Publishing, 2019.

Kundu

Central venous disease in hemodialysis patients: prevalence, etiology and treatment. J Vasc Access 2010; 11: 1–7.

Glanz

Gordon

Butt

KMH

, et al. The role of percutaneous angioplasty in the management of chronic hemodialysis fistulas. Ann Surg 1987; 206: 777–781.

Guo

Shi

Xie

, et al. Left innominate vein stenosis in an asymptomatic population: a retrospective analysis of 212 cases. Eur J Med Res 2017; 22: 3.

Salik

Daftary

Tal

MG.

Three-dimensional anatomy of the left central veins: implications for dialysis catheter placement. J Vasc Interv Radiol 2007; 18: 361–364.

10.

Rosenthal

RL.

The 50% stenosis. Am J Cardiol 2015; 115: 1162–1165.

11.

Oguzkurt

Tercan

Yıldırım

, et al. Central venous stenosis in haemodialysis patients without a previous history of catheter placement. Eur J Radiol 2005; 55: 237–242.

12.

Shi

Cheng

Song

, et al. Anatomical factors associated with left innominate vein stenosis in hemodialysis patients. Hemodial Int 2014; 18: 793–798.

13.

Kotoda

Akimoto

Kato

, et al. Central venous stenosis among hemodialysis patients is often not associated with previous central venous catheters. ASAIO J 2011; 57: 439–443.

14.

Oginosawa

Abe

Nakashima

Prevalence of venous anatomic variants and occlusion among patients undergoing implantation of transvenous leads. Pacing Clin Electrophysiol 2005; 28: 425–428.

15.

Song

Seo

Kang

, et al. Innominate vein stenosis in breast cancer patients after totally implantable venous access port placement. J Vasc Access 2015; 16: 315–320.

16.

Smith

Moureau

Vaughn

, et al. Patterns and predictors of peripherally inserted central catheter occlusion: the 3P-O study. J Vasc Interv Radiol 2017; 28: 749–756.

17.

Paquet

Boucher

Valenti

, et al. Impact of arm selection on the incidence of PICC complications: results of a randomized controlled trial. J Vasc Access 2017; 18: 408–414.

18.

Ming

Yuan

, et al. Effectiveness of catheter directed thrombolysis and stent implantation on iliofemoral vein thrombosis caused by iliac vein compression. J Thromb Thrombolysis 2017; 44: 254–260.

19.

Limesh

Desai

Kedalaya

, et al. A case of chylothorax in a hemodialysis patient with left innominate venous stenosis. Indian J Nephrol 2017; 27: 234–236.

20.

Sobrinho

Aguiar

Stent placement for the treatment of malignant superior vena cava syndrome—a single-center series of 56 patients. Arch Bronconeumol 2014; 50: 135–40.

21.

Haage

Vorwerk

Piroth

, et al. Treatment of hemodialysis-related central venous stenosis or occlusion: results of primary Wallstent placement and follow-up in 50 patients. Radiology 1999; 212: 175–180.

22.

Bakken

Protack

Saad

, et al. Long-term outcomes of primary angioplasty and primary stenting of central venous stenosis in hemodialysis patients. J Vasc Surg 2007; 45: 776–783.

23.

Ozyer

Harman

Yildirim

, et al. Long-term results of angioplasty and stent placement for treatment of central venous obstruction in 126 hemodialysis patients: a 10-year single-center experience. Am J Roentgenol 2009; 193: 1672–1679.