Abstract
Purpose
To determine the anatomical variations and morphology of the external carotid artery (ECA) and its anterior branches.
Methods
Using computed tomography angiography (CTA), the origin, internal diameter, and surface laterality emergence of the superior thyroid (STA), lingual (LA), and facial (FA) arteries were evaluated retrospectively evaluated and classified. The bifurcation level of the common carotid artery (CCA) in relation to the cervical vertebrae and disc was also determined.
Results
A total of 76 CTA were included in the study. STA originated from the carotid bifurcation (CB) (type I), CCA (type II) and ECA (type III) in 20.4 (31/152), 17.1 (26/152) and 50.7% (77/152) cases, respectively. Also 10.5% (16/152) arose from a shared trunk with LA as a thyrolingual trunk (TLT) (type IVa), and absent in 1.3% (2/152). LA originated in the CB in only one case. A linguofacial trunk (LFT) was present in 14.5% (22/152). Mean diameters of STA, LA and FA were 1.70, 1.95 and 2.45 mm, respectively. Meanwhile, surface laterality were predominately from anteromedial, medial, and anterior, respectively. CB was mainly on C3 or C3–C4 (55.9% of cases).
Conclusions
STA origin below the ECA is a common finding. Our population presented the highest percentage of TLT (10.5%) and high CB (9.8%) in literature. Considering these variations are important to prevent complications in neck surgical procedures.
Similar content being viewed by others
Explore related subjects
Discover the latest articles, news and stories from top researchers in related subjects.Avoid common mistakes on your manuscript.
Introduction
The external carotid artery (ECA) has three anterior branches, classically described: superior thyroid artery (STA), lingual artery (LA) and facial artery (FA). Several studies have reported variations in their origin, within the carotid triangle [11, 20, 24]. The typical STA origin from ECA has been reported with prevalence as low as 24–25% in Turkish and Spanish populations [29, 35, 40], and as high as 80.4–88.3–72% in Kenyan and Brazilian [27, 32].
Common origins have been described such as a thyrolingual trunk (TLT) [4, 16, 25], a linguofacial trunk (LFT), and a thyrolinguofacial trunk (TLFT) [23, 27]. Cappabianca et al. reported a prevalence in an Italian population of 10% of LFT, and 3.6% of TLT [4]. Natsis et al. found a higher prevalence of LFT (21%) and a similar one of TLT (3%) in Greeks [24].
There are several different classifications to describe STA origin variations. The classification proposed by Vazquez et al. is the most cited in the literature. Type I was defined when STA originated at the carotid bifurcation (CB) level, type II at the common carotid artery (CCA), type III at the ECA, and type IV when STA arose as a common trunk with 1 or more of the anterior ECA branches. The latter was then subdivided into IVa (TLT) and IVb (TLFT) [36]. Natsis et al. on the other hand, proposed another classification in which type I was for an independent STA which could arise from the ECA, CB, or CCA, and type II was for common trunks [24].
These branches have also been subject to anatomical studies in cadaveric specimens [9, 11, 20, 21, 24, 27,28,29, 31, 32, 35, 37]. Imaging studies such as the computed tomography angiography (CTA) with 3D reconstruction, is considered a high sensitivity head and neck study for the assessment of vascular structures [1, 4, 14, 33]. It allows better circulatory anatomic visualization, thus improving head and neck surgery planning or minimally invasive procedures (i.e. laparoscopic thyroidectomy, embolizations, selective chemoembolization, endarterectomy, fasciocutaneous flaps of face and neck, etc.) [3, 9, 10, 22, 28, 33, 36, 38]. Anatomical variations also need to be considered to avoid complications such as perforation and rupture that may cause hemorrhage, and the treatment of these [12, 26, 39].
There is a high prevalence of variants of the anterior ECA branch origin, with variability between populations [24, 27, 35]. However, data from Latino populations is limited [11, 32]. It is important to consider these during head and neck surgical procedures [10, 13, 15, 31]. The objective of our study was to determine the variations and morphometries of these branches.
Materials and methods
An observational, retrospective, transversal and descriptive study was performed. Imaging studies (CTAs) were obtained from the database of the Radiology and Image Department of the University Hospital “Dr. José Eleuterio González”. CTA examinations were performed with a 64-slices CTA scanner (General Electric CT99 Light Speed VCT, Software 2978195VCT). CTA parameters were: rotation 0.4 s helicoidal acquisition, 20 mm detector covering, 120 kV, 400+, 0.625 mm width slices, 0.53:1 mm/rot Pitch and 22 to 23 cm FOV.
A total of 152 ECA were included from 76 patients, of which 66% (50/76) were men and 34% (26/76) women, with a mean age of 58.7 ± 16.4 years (18–90 years). CTAs from adult patients (>18 years), without gender distinction, with adequate artery visualization were included. Those with a history of neck surgeries, structural alterations, great tumors, artifacts or cervical trauma were excluded. Those with low image quality or unidentifiable arteries were eliminated. The sample size was previously calculated based on the variability reported in the literature and a confidence of 95%. This resulted in a sample size of 114 carotid arteries however, it was decided to expand the sample size to 152.
Two observers reviewed and analyzed the image studies. Origin related to the carotid bifurcation and vertebral level, internal diameter, and surface laterality emergence of each anterior ECA branches (STA, LA, and FA) were recorded on a database. STA origin was classified according to Vazquez et al. 2009 classification (Fig. 1). CB origin was determined when the vessel’s emergence was in contact with a perpendicular line traced in the division of the CCA. The length was measured when a trunk between two (TLT, LFT) or more arteries was found (TLFT).
Statistical analysis was performed by 20.0 version SPSS computational program for Windows XP. Central tendency tests (average, standard deviation, frequency) were performed to analyze data. Chi-square non-parametric tests were made to assess significant differences between groups according to gender. p value equal to or lower than 0.05 was considered statistically significant.
This study was previously reviewed and approved by the University’s ethics and research committees under the registration number AH17-0001.
Results
STA had a typical origin in 50.7% (77/152) of the cases, originating as an only branch from the ECA, with another 37.5% (57/152) originating at a lower level (CB or CCA) (Fig. 1). The remaining 10.5% (16/152) of STAs originated from a common trunk (Table 1) with 1.3% (2/152) absent. Overall, including trunks, only 55.9% (85/152) originated from the ECA, 24.3% (37/152) from the CB, and 18.4% (28/152) from the CCA. STA originated directly from the carotidal system as an individual artery in 88.2% (134/152) of the cases. Only in one case, was the LA originating from the CB.
The origin of the STA from the carotid artery, concerning laterality, had a statistically significant difference (p 0.037), as well as in relation to gender (p 0.013) (Table 1).
The FA had the largest mean internal diameter and was the furthest from the CB with 2.45 ± 0.46 mm and 20.89 ± 7.66 mm, respectively. STA had a mean distance of 6.7 mm cephalic to the CB when originating from the ECA, and 7.5 mm caudally when originating from the CCA (Table 2).
The most common emerging surfaces were anterior and anteromedial for the STA and FA, while the LA was most prevalent on the anteromedial and medial surfaces. Only 2.6% (4/152) of LA and 2.0% (3/152) of FA originated from a posterior surface (Table 3).
TLT was found in 10.5% (16/152), emerging 8 from ECA and 6 from CB (Fig. 2) (Table 1). LA originated with FA, forming a LFT in 14.5% (22/152) (Fig. 3) (Table 1). The mean length of TLT was 6.60 mm and an internal diameter of 2.97 mm. For LFT, the mean length was 9.23 mm and an internal diameter of 3.10 mm. TLFT were not identified emerging from ECA nor CB in this sample population.
CB was most frequent in C3/C4 level with 29.6% (45/152), with overall, 79.6% (121/152) of CB occurring at the levels of C3, C3/C4 and C4 (Table 4).
Discussion
The STA is the most variable anterior branch of the ECA and is of great importance during the planning of any neck surgery, such as thyroidectomy, cricothyroidotomy, tracheostomy, laryngectomy, super-selective intra-arterial chemotherapy, endarterectomy, and minimally invasive procedures of the neck due to the risk of damage [1, 3, 4, 10, 12, 14, 15, 22, 26, 28, 33, 36, 38, 39]. Imaging studies do not classically report vascular variations; therefore, the surgeon should keep in mind the high anatomical variability of these to prevent complications, such as perforation and rupture [3, 5, 10, 12, 15, 26, 39]. Most STA studies have been performed in cadavers and few with image studies (Table 5).
STA origin has been a controversial topic. Embryologic studies confirm an origin from the CB and CCA, forming from them third aortic arch, while ECA derives from the first and a contribution of the second. Consequently, STA classification can be divided into two categories: one from CB and CCA and the other from just ECA [21, 24, 35].
Classically CCA has no collateral branches [2, 22]. STA origin incidence from the CCA have been reported from 3 to 45% [1, 4, 6, 8, 21, 24, 29, 32, 35, 37]. We observed an overall incidence of 17.1% (26/152), with a significantly higher prevalence in women (30.8% 16/52) than in men (12.0% 12/100) (p 0.013). At a CB level, prevalence ranges between 2 and 49%[4, 18, 19, 21, 24, 27, 35, 40]. We report a 20.4% (31/152) prevalence as an individual branch.
The common trunk of the anterior branches of the ECA variates. The TLT is composed of an artery or common trunk that after a short distance, branches the STA and LA. The trunk can originate from the ECA, CB or CCA. The latter is the least frequent [16, 25]. Past studies have reported a low incidence of TLT ranging from 0 to 3.5% [24, 25, 27, 31, 35]. In this study we found a high TLT prevalence in 10.5% (16/152) of the cases, including a bilateral case from CB, the highest reported in the literature, except for Won et al. [37], who reported a prevalence of 16.7% in 15 cadavers, without differentiating between TLT and TLFT.
The lingual and facial arteries have also been reported absent (agenesis) from the carotid artery, or as ectopic branches, such as an intra-parotid origin [15, 17, 19, 29, 31, 34]. An origin as a common trunk, such as the LFT have been reported between 7.5 and 31% [9, 21, 24,, 31, 40]. We report a 14.5% (22/152) prevalence, similar to that of a Turkish (7.5%) [29] and Italian (10%) [4] populations. TLFT are rare, most studies reporting below 1% [4, 11, 21, 23, 24, 31, 35], except for Ongeti et al. [27], who reported a 6.5% prevalence in a Kenyan study. We found no cases in our study.
STA was absent in 1.3% (2/152) cases, both on the right side. Therefore, 74 right and 76 left STA were analyzed. Laterality was significant, as 39 out of 65 (60%) samples with an abnormal origin (CB or CCA) were left-sided (p = 0.037); these results coincide with previous authors that attribute left preference due to the aortic arch evolutionary and anatomic asymmetry, as well as to the thyroid embryonic gen [21, 24, 35, 40].
CB is usually described at the same level as the superior border of the thyroid cartilage (C3–C4 vertebral level) [13, 15, 22, 37]. Zümre et al. studied 40 fetuses and reported CCA bifurcation from C3, C4, and C5 in 55, 35, and 10%, respectively from the right sides; while from C3 and C4 in 60 and 40%, respectively from left sides [40]. Similarly, Klosek et al. reported CB with 2.7% in C2/C3, 10.4% in C3, 20.9% in C3/C4, 30.2% in C4, 16.3% in C4/C5 and 8.1% in evel C5 [13]. This study found the most frequent CB was localized at the intervertebral disk of C3/C4 with 29.6% (45/152), with 79.6% (121/152) between the vertebral bodies C3 and C4. However, 9.8% (15/152) had a high bifurcation (above vertebral body C3) and 10.5 (16/152) a low bifurcation (below vertebral body C4). High bifurcation is rare [7, 13]. This presentation should warn a possible lesion to the hypoglossal nerve, a nerve that passes with the carotid artery and the jugular vein in the carotid sheath, as well as other risks during surgical procedures, such as minimally invasive surgery, catheterization, endarterectomy, fasciocutaneous flaps, etc [5, 7, 9, 10, 13, 15, 17, 19, 22, 29, 37].
Variations in the vascular anatomy have been correlated with a higher prevalence of complications. These should be considered during surgical procedures (conventional and minimally invasive) as they may be damaged or occluded due to their abnormal origin. They are also important during the treatment of complications such as unexpected hemorrhages. Common trunks should also be kept in mind during chemoembolization as treatment for malignancies, due to the shared origin and extensive blood supply [3, 9, 10, 12, 13, 15, 19, 20, 22, 23, 26, 29, 32,33,34, 40].
Narayanan et al. proposes that altered blood flow due to a high CB could originate a disproportionated growth and thus changes in ECA branches' origins [23]. In our study, there was no statistical relationship between a high CB and a higher incidence of STA from CB or CCA (p = 0.308).
Our study has several limitations. Although the sample was obtained from a large reference center of the northeastern part of Mexico, its results cannot be generalized, due to ethnical differences between the regions of the country. Measurements were performed by two observers. However, kappa coefficient for interobserver reliability was not performed. Strengths of this study include the calculation of a sample population, to validate results. Few studies include as many parameters in their analysis (STA origin, laterality, and gender correlation, morphometry of all anterior branches, the prevalence of individual branches and common trunks, surface of emergence, and vertebral level of bifurcation stratified by gender and laterality). A simple review was performed to compare results amongst populations and studies.
Conclusion
The STA originated from the ECA (Type III) as an independent artery, as described in the literature, in only half of the cases (50.7%, 77/152). A high incidence of variability was shown, with 37.5% (57/152) emerging from the CB (type I) or CCA (type II), and the remaining 11.8% originating from a common trunk (Type IVa, 16/152) or absent (2/152).
Our population presented a high percentage of TLT (10.5%, (16/152)), the highest in literature according to our knowledge. The percentage of TLFT (Type IVb) was null, similar to the low prevalence of other authors. STA origin variations were significantly different in gender (p 0.013), with more men having an origin at the CB (28.0 [28/100) vs 17.3% [9/52]), and more women at the CCA (30.8 [16/51] vs 12.0% [12/100]). CCA origin was also statistically (p 0.037) more common on the left side than the right (26.3 [20/76] vs 10.5% [8/76]).
We propose considering a high prevalence of anatomical variations in our population, in regards to the anterior branches of the ECA, to prevent complications in surgical procedures (i.e. conventional open neck surgery of the thyroid, trachea, larynx and adjacent structures, minimally invasive procedures with implementation of laparoscopy, endoscopy and robotic surgery of the neck, as well as embolization, and selective chemoembolization, treating complications such as hemorrhages, etc.) of our population.
References
Acar M, Salbacak A, Sakarya ME, Zararsiz I, Ulusoy M (2013) The morphometrical analysis of the external carotid artery and its branches with multidetector computerized tomography angiography technique. Int J Morphol 31(4):1407–1414. https://doi.org/10.4067/S0717-95022013000400042
Adachi B (1928) Das Arteriensystem der Japaner, vol 2. Maruzen, Kyoto, pp 196–242
Bliss RD, Gauger PG, Delbridge LW (2000) Surgeon's approach to the thyroid gland: surgical anatomy and the importance of technique. W J Surg 24:891–897. https://doi.org/10.1007/s002680010173
Cappabianca S, Scuotto A, Iaselli F (2012) Computed tomography and magnetic resonance angiography in the evaluation of aberrant origin of the external carotid artery branches. Surg Radiol Anat 34:393–399. https://doi.org/10.1007/s00276-011-0926-3
Dessie MA (2018) Variations of the origin of superior thyroid artery and its relationship with the external branch of superior laryngeal nerve. PLoS ONE 13(5):e0197075. https://doi.org/10.1371/journal.pone.0197075
Esen K, Ozgur A, Balci Y, Tok S, Kara E (2017) Variations in the origins of the thyroid arteries on CT angiography. Jpn J Radiol 36(2):96–102. https://doi.org/10.1007/s11604-017-0710-3
Gluncic V, Petanjek Z, Marusic A, Gluncic I (2001) High bifurcation of common carotid artery, anomalous origin of ascending pharyngeal artery and anomalous branching pattern of external carotid artery. Surg Radiol Anat 23(2):123–125. https://doi.org/10.1007/s00276-001-0123-x
Gupta P, Bhalla AS, Thulkar S (2014) Variations in superior thyroid artery: a selective angiographic study. Indian J Radiol Imaging 24(1):66–71. https://doi.org/10.4103/0971-3026.130701
Hayashi N, Hori E, Ohtani Y (2005) Surgical anatomy of the cervical carotid artery for carotid endarterectomy. Neurol Med Chir 45:25–30
Inabnet WB III, Jacob BP, Gagner M (2003) Minimally invasive endoscopic thyroidectomy by a cervical approach. Surg Endosc 17:1808–1811. https://doi.org/10.1007/s00464-002-8760-7
Itezerote AM, Rodrigues CFS, Prates JC (1999) Arteria tiroidea superior: origen, ramas colaterales y glandulares. Rev Chil Anat 17(1):47–50. https://doi.org/10.4067/S0716-98681999000100007
Jeganath V, McElwaine JG, Stewart P (2001) Ruptured superior thyroid artery from central vein cannulation: treatment by coil embolization. Brit J Anaesth 87(2):302–305. https://doi.org/10.1093/bja/87.2.302
Klosek SK, Rungruang T (2008) Topography of carotid bifurcation: considerations for neck examination. Surg Radiol Anat 30(5):383–387. https://doi.org/10.1007/s00276-008-0337-2
Kramer M, Vairaktaris E, Nkenke E, Schlegel KA, Neukam FW, Lell M (2008) Vascular mapping of head and neck: computed tomography angiography versus digital subtraction angiography. J Oral Maxillofac Surg 66(2):302–307. https://doi.org/10.1016/j.joms.2007.05.031
Krmpotic-Nemanic J, Draf W, Helms J (2012) Surgical anatomy of head and neck. Springer Science and Business Media, Berlin
Lemaire V, Jacquemin G, Medot M, Fissette J (2001) Thyrolingual trunk arising from common carotid artery—a case report. Surg Radiol Anat 23:135–137. https://doi.org/10.7860/JCDR/2017/23144.10365
Lippert H, Pabst R (1985) Arterial variations in man. JF Bergmann Verlag, München, pp 62–65
Lo A (2006) Anatomical variations of the common carotid bifurcation. ANZ J Surg 76:970–972. https://doi.org/10.1111/j.1445-2197.2006.03913.x
Lucev N, Bobinac D, Maric I, Drescik I (2000) Variations of the great arteries in the carotid triangle. Otolaryngol Head Neck Surg 122(4):590–591. https://doi.org/10.1067/mhn.2000.97982
Manjunath CS, Lokanathan TH (2016) Anatomical variations in the origin of superior thyroid artery and its clinical significance. Int J Anat Res 4(3):2656–2658. https://doi.org/10.16965/ijar.2016.295
Mata JR, Mata FR, Souza MCR, Nishijo H, Ferreira TAA (2012) Arrangement and prevalence of branches in the external carotid artery in humans. Ital J Anat Embryol 117(2):65–74
Michalinos A, Chatzimarkos M, Arkadopoulos N, Safioleas M, Troupis T (2016) Anatomical considerations on surgical anatomy of the carotid bifurcation. Anat Res Int 2016:6907472. https://doi.org/10.1155/2016/6907472
Narayanan S, Murugan S (2018) Association of high carotid bifurcation and thyrolinguofacial trunk: a rare variation. Anat Sci Int 93(2):307–310. https://doi.org/10.1007/s12565-017-0405-y
Natsis K, Raikos A, Foundos I (2011) Superior thyroid artery origin in Caucasian Greeks : a new classification proposal and review of the literature. Clin Anat 24(6):699–705. https://doi.org/10.1002/ca.21181
Nochikattil SK, Sajith Babu TP, Manoharan S (2017) Thyrolingual trunk arising from common carotid artery—a case report. J Clin Diagn Res 11(8):AD01–AD02. https://doi.org/10.7860/JCDR/2017/23144.10365
Nomura M, Mori K, Seki S (2015) Perforation of superior thyroid artery by guidewire during carotid artery stenting. Neuroradiol J 28(3):329–332. https://doi.org/10.1177/1971400915594928
OngetiOgeng’O KWJA (2011) Variant origin of the superior thyroid artery in a Kenyan population. Clin Anat 25(2):198–202. https://doi.org/10.1002/ca.21207
Ozgur Z, Govsa F, Celik S, Ozgur T (2009) Clinically relevant variations of the superior thyroid artery: an anatomic guide for surgical neck dissection. Surg Radiol Anat 31(3):151. https://doi.org/10.1007/s00276-008-0405-7
Ozgur Z, Govsa F, Ozgur T (2008) Anatomic study assessment of origin characteristics of the front branches of the external carotid artery. J Craniofac Surg 19(4):1159–1166
Quain R (1844) Anatomy of the arteries of the human body. Taylor & Walton, London, pp 538–543
Shintani S, Terakado N, Alcalde R, Tomizawa K, Nakayama S, Ueyama Y, Ichikawa H, Sugimoto T, Matsumura T (1999) An anatomical study of the arteries for intraarterial chemotherapy of head and neck cancer. Int J Clin Oncol 4:327–330. https://doi.org/10.1007/s101470050079
Sreedhan R, Krishna L, Shetty A (2018) Origin of superior thyroid artery: under the surgeon’s knife. J Vasc Bras 17(4):290–295
Terayama N, Sanada J, Matsui O, Kobayashi S, Kawashima H, Yamashiro M et al (2006) Feeding artery of laryngeal and hypopharyngeal cancers: role of the superior thyroid artery in superselective intraarterial chemotherapy. Cardiovasc Inter Rad 29(4):536–543. https://doi.org/10.1007/s00270-005-0094-0
Tubbs SR, Shoja M, Loukas M (2016) Bergman's comprehensive encyclopedia of human anatomic variation, 1st edn. Wiley-Blackwell, Malden, p 522
Vázquez T, Cobiella R, Maranillo E, Valderrama FJ, Mchanwell S, Parkin I, San JR (2009) Anatomical variations of the superior thyroid and superior laryngeal arteries. Head Neck 31(8):1078–1085. https://doi.org/10.1002/hed
Wilson JL, Rozen WM, Ross R, Findlay MW, Ashton MW, Behan FC (2012) The superior thyroid artery perforator flap: anatomical study and clinical series. Plast Reconstr Surg 129(3):641–646. https://doi.org/10.1097/PRS.0b013e3182402eb2
Won S (2016) Anatomical considerations of the superior thyroid artery: its origins, variations, and position relative to the hyoid bone and thyroid cartilage. Anat Cell Biol 49:138–142
Yonenaga K, Tohnai I, Mitsudo K, Mori Y, Saijo H, Iwai T, Yonehara Y, Ota Y, Torigoe K, Takato T (2011) Anatomical study of the external carotid artery and its branches for administration of superselective intra-arterial chemotherapy via the superficial temporal artery. Int J Clin Oncol 16:654–659
Yu NH, Jahng TA, Kim CH, Chung CK (2010) Life-threatening late hemorrhage due to superior thyroid artery dissection after anterior cervical discectomy and fusion. Spine 35(15):E739–E742. https://doi.org/10.1097/BRS.0b013e3181cf46b4
Zümre O, Salbacak A, Ciçekcibaşi AE, Tuncer I, Seker M (2005) Investigation of the bifurcation level of the common carotid artery and variations of the branches of the external carotid artery in human fetuses. Ann Anat 187:361–369. https://doi.org/10.1016/j.aanat.2005.03.007
Author information
Authors and Affiliations
Contributions
MH-N: Idea development, Protocol/project development, Submission to Committees, Data collection and management, Data analysis, Manuscript writing, manuscript editing. JLM-G: Protocol/project development, Data collection and management, Data analysis, Manuscript writing. RP-R: Protocol/project development, Data collection and management, Data analysis, Manuscript writing. GE-R: Protocol/project development, Submission to Committees, Data analysis, manuscript editing. AQ-G: Idea development, Protocol/project development, Submission to Committees, Data collection and management, Data analysis, Manuscript writing, manuscript editing. BAF-R: Idea development, Protocol/project development, Data analysis, Manuscript writing. REE-O: Idea development, Protocol/project development, Data analysis, Manuscript editing. SG-L: Idea development, Protocol/project development, Data analysis, Manuscript editing.
Corresponding author
Ethics declarations
Conflict of interest
This study was not funded by any academic or industry resources. Therefore, authors declare the absence of conflict of interest.
Ethical approval
This study was submitted and approved by the local research committee.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
About this article
Cite this article
Herrera-Núñez, M., Menchaca-Gutiérrez, J.L., Pinales-Razo, R. et al. Origin variations of the superior thyroid, lingual, and facial arteries: a computed tomography angiography study. Surg Radiol Anat 42, 1085–1093 (2020). https://doi.org/10.1007/s00276-020-02507-6
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00276-020-02507-6