The anatomical and physiological background for such a hypothesis started more than a century ago when the medulla oblongata was recognized vital for cardiovascular function after Dittmar in 1873 observed a dramatic fall of blood pressure by transection of the brainstem caudal to the pontomedullary junction [18]. Later on, in 1946, the crucial role in blood pressure regulation of the sympathetic centres of the medulla was evidenced by Alexander who succeeded in obtaining a tonic effect on blood pressure by stimulating the medullary ventrolateral area [2]. Then, in the 1960s and 1970s, the presence of monoaminergic neurons was identified in the brainstem [15, 24], as well as the existence of bulbospinal projections to the intermediolateral cell column [3]. Soon after, in the 1980s, the localization in the RVL medulla of vasopressor neurons involved in arterial pressure control, namely, cell bodies containing adrenaline/noradrenaline, was demonstrated [4, 6, 16, 26, 43, 44]. Also, strong links were evidenced from the RVL medulla to the nucleus tractus solitarii [11, 45], as well as direct projections of the cardiovascular inputs entering the VL medulla onto the central autonomic area of the thoracic cord [8, 9].

Lesioning of the epinephrine neurons in the VL medulla was shown abolishing the vasodepressor components of the baroreflex and the cardiopulmonary reflex [22]. Meanwhile, experiments in baboons were performed by the group of Jannetta to induce arterial hypertension by mimicking neurovascular pulsating compression in the VL medulla; authors used a pulsatile inflated balloon device [48–50]. Haemodynamic changes were observed after pulsatile compression is performed on the left side, but not on the right side [49].

The first clinical report on vascular compression as an aetiology of essential HT was by Jannetta and Gendell in 1979 [27]. They observed a high rate of neurovascular compression of the VL medulla on the left side in their patients affected with trigeminal neuralgia or hemifacial spasm when there was an associated hypertension, compared to the patients who did not have HT. Since then only two surgical studies were published: one retrospective, by Jannetta et al. [28], and one prospective, by Geiger et al. [20]. In Jannetta et al’s series, of the 53 patients with essential HT, 51 had vascular compression at the left VL medulla. Among the 36 patients who benefited from a decompression considered “adequate”, 32 (i.e. 89 %) had subsequent blood pressure normalization. In Geiger et al’s series, of the 10 patients followed from 5 to 66 months, 8 (80 %) were improved, 3 of them requiring no further antihypertensive medications. Improvement in blood pressure occurred more often in the patients who underwent surgery on the left side [20, 28, 33, 39]. The predominant left-sided lateralization of blood pressure control is concordant with the prevalence of the left IX–Xth nerve complex in conveying the afferences originating from the baroreceptors of the (left-sided) cardiac atrium.

Because all patients affected with essential HT not medically controlled cannot be candidates for surgical operation, a large number of studies were launched to evaluate the validity of MRI exploration to find out neurovascular conflicts (NVC) susceptible to cause the disease. Over the past 10 years, as many as 15 studies were carried out to assess the eventual role of a vascular compression at the brainstem in the genesis of HT [1, 10, 12, 21, 25, 29, 36–38, 40, 46, 51, 56, 58, 59, 62]. Of the 12 publications with detailed information, a majority (nine) reported a higher rate of images of vascular contact/compression (vc/c) at the VL medulla in the group of patients with essential HT, compared to the group of patients with normal blood pressure [1, 21, 25, 29, 36–38, 46, 51]. Also, of the ten publications providing information on the side of the compression, in six a higher rate of left-sided vc/c over right-sided vc/c was found in the essential HT group [1, 37, 38, 40, 46, 51, 58]. Further, in a recently published meta-analysis, including 597 patients with apparent essential hypertension and 609 controls, left-sided NVC at the VL medulla was prevalent in patients with essential HT. However, there was no significance when subanalysis was confined to the prospective studies – p = 0.178 for prospective vs. p = 0.001 for retrospective studies [7]. Thus, predictive value of MR imaging for selecting patients remains uncertain.

A major limitation for MRI screening is the difficulty to discriminate whether a detected elongated arterial loop in relation with the VL medulla and the adjacent IXth–Xth REZ is the cause or the consequence of the raised blood pressure. To address this problem several authors compared imaging findings between patients with apparent primary HT and patients with secondary HT [1, 21, 36, 37, 40, 46, 58].Studies showed that in the group of patients with secondary HT, rate of NVC at the brainstem was less than in the group of patients with primary, i.e. essential, HT (p = 0.01). In addition when an image of the megadolicho-artery was present, the vascular image was not predominantly left sided. However, these studies did not allow any definitive conclusion, due to the fact that the number of patients enrolled was not sufficient to reach statistical significance [7].

Lack of reliability of MR screening to ascertain responsibility of neurovascular images in the genesis of HT makes it difficult to indicate vascular decompressive surgery in patients solely affected with apparent essential HT. Therefore, it has been considered wise to proceed through indirect ways to investigate the potential effect of MVD on pharmacoresistant HT, namely, investigating patients referred for HFS and having associated HT, as initially explored by Jannetta and Gendell [27].

In an ongoing study, we are assessing the effects obtained on blood pressure (BP) by vascular decompression of the IX–Xth REZ and adjacent VL aspect of the medulla (in addition to MVD of the facial REZ) in a group of 48 patients operated on for their hemifacial spasm (HSF) and who presented with an associated apparent essential HT. These 48 patients represented 23.88 % of the 201 patients who were referred for surgery of their HFS, over the past 20 years. Long-term effect of the MVD on patients’ HT is at present under investigation. At first look, effect is promising; once statistical study will be completed, results will be submitted for publication.

Decompression of the IXth–Xth REZ and adjacent VL medulla, in addition to decompression of the facial REZ, does not require much more extensive surgical approach than the classical one for HFS [5, 35, 42, 47, 52, 53]. High-resolution MR imaging with special sequences allows to identify NVC with very high sensibility and specificity [31, 32, 54, 55].

Whether such an indirect approach is valuable to bring insight to knowledge must be questioned at first. Is the percentage of patients affected with HT associated with HFS higher than the one in the general population of same range of age? Whilst some publications report a higher prevalence of HT in patients with HFS than in control groups [17, 41], as high as in 67 % in the later publication, others did not find any statistically significant differences: Colosimo et al., 47.3 % vs. 52.7 % [13, 14], and Tan et al., 42.7 % vs. 39.1 % [57]. Discrepancy between these studies might be bias in the publications; studies with prevalence were of retrospective nature and/or were lacking of proper matching with controls.

An important study is the one by Nakamura et al., which brings convincing arguments for the role of vascular compression in the genesis of HT in patients with left-sided HFS harbouring NVC image at MRI exploration [38]. In this study there was no significant difference in prevalence of HT between the patients with HFS (82 patients enrolled) and the control group without HFS (similarly 82 patients enrolled, with age and sex matched), 39 % vs. 29.3 % (p = 0.19); there was also no difference between the patients with left-sided (n = 44) and right-sided (n = 38) HFS, 32 % vs. 47 %. Conversely, difference was significant when side and presence or absence of HT were considered together. As a matter of fact, a vascular compression at the VL medulla was observed at MRI in 86 % of the patients with left HFS who had HT (n = 14) vs. in only 33 % of the patients with left HFS but without HT (n = 30) (p = 0.0012). To be mentioned, there was no such significance in patients with right HFS (p = 0.18). Authors concluded that these findings are of clinical importance, at least for the patients who harbour left HFS and HT and perhaps also for those with sole essential HT provided that MRI shows vascular compression at the brainstem.