Cross-sectional area variations of internal jugular veins during supine head rotation in multiple sclerosis patients with chronic cerebrospinal venous insufficiency: a prospective diagnostic controlled study with duplex ultrasound investigation

Chronic cerebrospinal venous insufficiency (CCSVI) is a congenital syndrome affecting the extracranial vessels (internal jugular and azygos veins) that is characterised by different valve malformations, stenoses and segmental or global hypoplasia with impaired venous drainage, opening collateral circulation [1‐3]. Several studies have demonstrated a strong association between multiple sclerosis (MS) and CCSVI, but subsequent clinical research has failed to support this hypothesis [4‐13]. Thus, the debate currently continues over whether the relationship is real. In fact, some reports have claimed that the data in the literature have been insufficient to establish the importance of CCSVI as a major factor in MS pathogenesis. Echo-colour Doppler (ECD) studies of CCSVI are usually performed according to the five Zamboni criteria; the detection of at least two of these parameters indicates a diagnosis of CCSVI. Nevertheless, ECD investigation is still not currently standardised in many respects, and it is overly dependent on individual patients and operators, with high interobserver variability for untrained examiners [14, 15]. However, different studies have reported that ECD is more sensitive than the magnetic resonance venography (MRV) in detecting intraluminal jugular defects, while MRV is more sensitive in showing collaterals [9, 16]. Both techniques have proved effective in assessing the size and course of venous vessels of the neck, showing internal jugular vein (IJV) asymmetry in both normal subjects and MS patients. The left internal jugular vein is usually smaller than the right due to preferential intracranial venous drainage through one sigmoid sinus versus the other, with weak correlations with age and gender [17‐21]. Venous size can vary depending on hydration status, position, cardiac status, thoracic pump, head position, and compression from adjacent structures [22‐26]. Even dysfunction of the cardiovascular autonomic nervous system can reduce vascular tone, affecting jugular size [27]. More frequently, MS patients have shown a jugular cross-sectional area (CSA) ≤30 mm² compared to healthy controls (HCs) (43.5% vs. 16.7%) [28]. These ultrasonographic findings were also confirmed by catheter venography (CV) and were mainly detected in the middle part of the IJV, at the level of the cricoid cartilage immediately below the sternocleidomastoid muscle (SCM) [29]. They have not been clearly correlated with intraluminal defects, although they have been explained by low values of internal pressure [30]. At this level, CV and ECD have shown a collapsed jugular vein with loss of its elliptical appearance under SCM imprinting [5, 29]. Patients have usually been examined in the supine and sitting positions, in accordance with a static protocol without head rotation (0° from midline). A dynamic approach to assess venous sizes with different degrees of head rotation has only been applied to improve jugular venous catheterisation. These echographic studies have suggested that head rotation to the contralateral side increases the CSA of IJVs in supine subjects [25, 31]. The aim of this study was to evaluate the behaviour of the CSA of the IJVs during supine head rotation in MS patients with CCSVI, compared to HCs.

A total of 1,222 IJVs were evaluated: 596 in the HCs group and 626 in the MS group. Table 2 shows the distribution of the five Zamboni criteria in the two groups. As reported in the literature, our reconstruction of the longitudinal ultrasound scans of IJVs in MS patients detected normal veins (11.7%), veins with only VDs (49.2%) and hypoplastic veins (7.5%). We also identified two new morphological types of IJVs, which we defined as “miopragic” (27.1%) and “miopragic with VDs” (4.5%). A miopragic jugular vein has a typical aspect of an “hourglass”, with an extremely small diameter in the intermediate portion (mean ± SD = 3.0 ± 0.8 mm) and larger values at the J1 (10.5 ± 1.0 mm) and J3 levels (5.3 ± 0.6 mm) (Figure 5). In 4.5% of cases, miopragic veins presented with proximal VDs and mean diameter values that were comparable to previous types (J1 = 10.2 ± 1.1 mm; J2 = 3.3 ± 0.7 mm; J3 = 5.2 ± 0.5 mm) (Figure 6). The HCs showed normal jugular veins and proximal VDs in 93.5% and 6.5% of cases, respectively. In addition, we detected no miopragic or hypoplastic types in this subject group. Table 3 clearly describes these results, and as shown, the differences in frequency of IJV morphological types between the patients and controls were highly significant (Χ ² [4, N = 1222] = 825.129, p < 0.001). In a per IJV analysis, we compared the ΔCSA measurements after head rotation in the MS patients and HCs, with regard to their IJV morphological types. Significant differences between jugular CSAs before and after the reported manoeuvre were only observed in the MS patients (F[6,1215] = 6414.57, p < 0.001). The results are summarised in Table 4 and Figure 7. In a per patient analysis, we also correlated the IJV morphological types with the CCSVI scores (Table 5). A strong association was observed between the IJV morphological types and the CCSVI scores (Χ ² [9, N = 313] = 75.183, p < 0.001). In fact, the patients with isolated VDs had a median CCSVI score of two, those with hypoplasia plus VDs or miopragia plus VDs had a median score of three, and the patients with hypoplasia of an IJV and miopragia plus VDs of the other IJV had a median score of four. The above-mentioned classification of the complexity of the patients’ IJV morphological types was able to explain the variability of CCSVI scores in the MS group. In our series, we reported a statistically significant difference in the mean ranks of CCSVI scores among the different morphological types of IJVs (Kruskal-Wallis H[3] = 48.094, p < 0.001). The mean rank was 110.60 for isolated VDs, 153.59 for hypoplasia plus VDs, 177.93 for miopragia plus VDs and 219.43 for patients with hypoplasia of an IJV and miopragia plus VDs of the other IJV. Table 6 shows post-hoc paired comparisons of CCSVI scores among the IJV morphological types after Kruskal-Wallis analysis. The patients with hypoplasia + miopragia + VDs had higher CCSVI scores than patients with hypoplasia + VDs (H₃ hypothesis, p = 0.033), while the patients with hypoplasia + VDs had higher CCSVI scores than patients with only VDs (H₄ hypothesis, p = 0.024). Likewise, the patients with miopragia + VDs had higher CCSVI scores than patients with only VDs (H₅ hypothesis, p < 0.001), and the patients with hypoplasia + miopragia + VDs had higher CCSVI scores than patients with only VDs (H₆ hypothesis, p < 0.001). Wall miopragia was mainly observed in MS patients with the PP (70.0%) and SP (59.3%) clinical forms compared to the RR (48.3%) form (p = 0.015). In contrast, no significant differences in the observed frequencies of wall miopragia were noted among the four classes of disease duration (56.0%, 58.0%, 45.3%, 60.6%; p = 0.260); this result did not change even after distinguishing for clinical forms.

Currently, there is much ECD evidence regarding the presence of proximal jugular valve malformations with altered local flow in MS patients with different prevalence data according to the authors [1, 7, 15, 37]. In our experience we have found CCSVI in 89.8% of MS patients and in 5.4% of HCs. Similar results have been obtained with other diagnostic imaging techniques, such as MRV [9, 38, 39] and CV [8, 37]. However, all diagnostic methods have technical issues, and not all are completely standardised [9, 14]. Since 2009, jugular abnormalities have been included in the Consensus Document of the International Union of Phlebology (IUP) on venous malformations and have been classified as congenital truncular lesions [40], but until now, little was known regarding their histopathogenesis. The recent work of Coen was the only study that addressed the issue from this point of view, and it was the first that detected an altered ratio of type I/III collagen in the IJVs of MS patients, without any differences in cellularity or connective tissue distribution [41]. This condition has been similarly described in many other apparently unrelated conditions (e.g., varicose saphenous veins, haemorrhoids, paraoesophageal hernia, skin incisional hernia, recurring inguinal hernia, pelvic organ prolapse), suggesting connective tissue systemic involvement [42‐44]. These structural changes were also found in vessel wall samples taken far from balloon angioplasty treatment areas, suggesting their independence on local trauma. As evidenced by several studies, type I collagen provides mainly tensile strength and rigidity by building thick fibre that offer resistance to tissue, while type III collagen determines elasticity [43]. These histochemical data could explain the typical hourglass appearance of a part of the IJVs that we defined as “miopragic”. This morphological type was only detected in MS patients in the supine neutral position, with related flow blocked in their middle portions. It could be the result of vein collapse under SCM pressure because of reduced wall stiffness (wall miopragia). In fact, in the resting position, muscle pressure is greater at the J2 level, where the SCM is larger and closer to the IJV, crossing it to pass medially. Here, the vein is usually positioned anterolateral to the carotid artery [45]. Similarly, external pressure (exerted on the vein) is less in the distal segment (J3), where the SCM is located lateral to the vessel, and also in the proximal portion (J1), where the jugular passes among the sternal and clavicular heads of the same muscle (Figure 8). In addition, the regular outflow along the IJV ensures its patency at the J3 level, with a shunt through the common facial vein towards the external or anterior jugular veins. The inferior thyroid veins mainly support J1 outflow. Contralateral rotation of the head restores regular flow along the vein through the significant increase (F[6,1215] = 6414.57, p < 0.001) in CSA in the intermediate section, which was only observed for this morphological type (see Additional files 1, 2 and 3) and was never detected in the HCs. This result in the HCs agrees with what Lorchirachoonkul and Suarez found in normal subjects, namely a non-significant increase in jugular CSAs and longitudinal diameters with contralateral rotation of the head [19, 25]. However, the jugular collapse could also be an expression of cardiovascular autonomic disorders, as described in MS patients by several authors [27, 46]. A dysregulation of vegetative function would reduce venous tone, causing jugular wall collapse under SCM pressure. However, this datum is controversial because older studies, as well as recent publications, have observed no anomalies on orthostatic tests [47, 48]. According to other researchers, autonomic disorders are present in up to 23% of MS patients [49], and this finding would not explain the jugular collapse that we found in a greater number of subjects (54.9%). Among other findings, our data revealed no statistically significant difference in the mean CSA of normal IJVs between MS subjects with CCSVI and HCs in the supine position, as shown in another study [28]. Similarly, the distribution of mean values of the CSA of normal jugular veins in HCs in the supine position corresponded to what has been reported by several authors [25, 28, 50], with a trend toward an increase when passing from J3 to J1 [20, 24] and with rotation of the head [19, 25]. Table 7 summarises these results, and as shown, our values of CSA are similar to those of other researchers, while they differ from those of Zamboni’s two studies. We believe that our higher values might be due to the younger mean age and likely lower body weight of the participants in the Zamboni’s study, compared to our and other studies. In fact, as proved by Mortensen, the internal jugular size has the best positive correlation with body weight [21]. As reported in a recent work, the failure of jugular angioplasty in one MS patient with CCSVI showed a totally collapsed middle part of the vein [26]. Surgical decompression of an atypical omohyoid muscle successfully solved this case. Others have described the omohyoid muscle as involved in the physiologic dilatation of the IJVs during opening of the mouth and deep inspiration, through short segmentary compression of the proximal part of the vein (J1) near the tendinous intersection [51]. This physiological status could become pathological in cases of prolonged compression, as detected in the lower part of the neck, or in the presence of a shorter omohyoid directly merged with the sternohyoid muscle. Patra instead was not able to locate precisely the position of the omohyoid muscle and to measure the jugular surface below it because of poor muscle echogenicity [52]. However, anatomic variations in the inferior belly of the omohyoid muscle have been rare, seen in only 3% of autopsy findings [53]. In our study, all extrinsic jugular compressions occurred in the middle part (J2) of the vein below the SCM and were therefore not attributable to the omohyoid muscle action. When investigating MS patients using MRV, some authors have found extrinsic severe stenoses in 22% of the patients’ IJVs [16]. These findings indicate that CCSVI is mainly caused by valvular malformations [1, 37] and that extrinsic compression by a muscle could represent a less frequent nonvalvular cause of compromised cerebral venous outflow [26, 54]. We believe that jugular collapse, secondary to extrinsic muscular compression, is the expression of a congenital weakness of vessel walls, likely because of a dysregulation of collagen synthesis. We support our assertions with the presence of miopragic jugular veins (combined with VDs) in 54.9% of MS patients and the absence of them in the HCs, for whom a low percentage of VD (6.5%) was detected instead. Therefore, if jugular valve defects can also be recognised in HCs, wall miopragia would seem to be an exclusive feature of MS patients. If we add to this percentage the 10.5% of patients with hypoplastic jugular veins and the 4.5% of patients with combined malformations (hypoplasia + miopragia + VDs), the total value of MS patients with wall miopragia and VDs reaches 69.9%, compared with 30.1% of those with VDs only. These data assume greater importance, given the close correlation of vein miopragia to the degree of haemodynamic impairment of cerebrospinal efferent vessels. In fact, as noted in our per patient analysis, jugular miopragia was mainly correlated with medium to severe scores (3–4) for CCSVI, while patients with isolated VDs show lower scores (2). Similarly, combined malformations are correlated with higher CCSVI scores. We stated that wall miopragia would not affect the totality of MS patients because in our study, we found jugular veins with isolated VDs in 30% of subjects, but this number could be higher than reported. In fact, the collagen distribution might not be the same from one patient to another or even between different segments of the same vein; therefore, we could have jugular vessels with altered collagen but not sufficiently altered to induce wall collapse. These veins would appear as normal vessels with isolated VDs. Moreover, our data demonstrated a greater prevalence of jugular wall miopragia in patients with the PP and SP clinical forms, compared to the RR form (p = 0.015), and no significant relationship between jugular wall miopragia and MS duration, supporting the hypothesis that this morphological type is congenital. The principal limit of this study was the inability to confirm our hypothesis of jugular wall miopragia by means of a direct histochemical analysis of the vessels.

The reconstruction of longitudinal echographic scans of IJVs provides a full view of vessels, thereby enabling the identification of a new morphological type with an hourglass appearance, which was not detected in the HCs. This type of vein shows a unique behaviour with the SCM stretching manoeuvre, most likely because of a condition of wall miopragia, the congenital nature of which was clearly shown by our study, but the meaning of which will have to be investigated further. This dynamic approach, applied to the conventional static ultrasound screening for CCSVI, allowed us to introduce the first selective criterion for angioplasty. In fact, it would be unthinkable to treat miopragic veins; balloon angioplasty would most likely fail because of the increased distensibility of the venous wall. Obviously, further histochemical studies will be needed to confirm whether jugular collapse, found in MS patients, could be the expression of dysregulation of collagen synthesis.