Introduction

Mid-aortic syndrome (MAS) is a rare condition presenting with severe hypertension (HTN) and characterized by severe narrowing of the abdominal aorta, usually involving the renal arteries and visceral branches [1]. The clinical characteristics of MAS were first described in 1963 [2], with the narrowing of the subisthmal aorta (unlike the typical Takayasu-type changes of the aortic arch and Leriche-type of aortic bifurcation obliterative disease) [3].

MAS has subsequently emerged as its own clinical entity and specific cause of renovascular hypertension (RVH) in childhood and young adulthood. Previously, MAS was termed coarctation [4–7], hypoplasia or mesenchymal disease of the abdominal aorta [8, 9], but the term MAS is the most recent preferred terminology as it provides an anatomical and clinical definition (as opposed to aetiopathogenetic disease entities) [10]. In most cases MAS is associated with renovascular disease, although, in previous studies among a total of 196 children with renovascular disease, 51 (26%) had MAS [11–14].

The aetiology of MAS is still unknown, and its pathophysiology remains elusive, although many mechanisms have been advocated. It has been described as a congenital disorder when diagnosed in neonates and associated with congenital syndromes [15–18]. However, Indian and Asian authors have described MAS as an acquired disorder [19, 20] when detected in adolescent and young adults with signs of perivascular and intravascular inflammation in conjunction with elevation of serum inflammatory markers. This form was described as an aorto-arteritis due to Takayasu disease.

MAS has been treated with increasing frequency over the past two decades, although many aspects of this syndrome still need to be clarified. Therefore, we conducted a retrospective review of the records of patients with hypertension related to MAS who have been followed in our centre. We describe the clinical syndrome, its multidisciplinary management, the response to our therapeutic approaches and the occurrence of renal dysfunction.

Materials and methods

Patients

Thirty-six children with hypertension and suspected renovascular disease who were found to have MAS with angiographic evidence of narrowing of the abdominal aorta were included in the study. They had been treated at Great Ormond Street Hospital for Children NHS Trust from November 1976 to July 2008. All children were diagnosed by digital subtraction angiography, which was considered the gold standard for suspected RVH [21]. A full medical examination, biochemical analysis and abdominal ultrasonography were performed as first approach in all the children. Less invasive imaging techniques were carried out on selected patients for whom the need for angiography was uncertain [21]. Two patients underwent computed tomography angiography, and another two children had magnetic resonance angiography. Abdominal-vessel Doppler ultrasound was not routinely performed because it was considered unhelpful in detecting stenosis of the smaller arteries [22]. All children were treated with anti-hypertensive agents as first attempt. Usually, they required a combination of different types of medications: beta-blockers, vasodilators, alpha and beta blockers. The angiotensin-converting-enzyme inhibitors and the angiotensin receptor-blockers were used as second choice, when other options failed to control blood pressure (BP), with close monitoring of the renal function.

The hospital records were reviewed, and data on initial presentation, treatment and follow-up were obtained and recorded.

The BP response to treatment was evaluated as:

  1. (A)

    cure: normalization of BP to below the 95th centile for age, gender and height centile without any anti-hypertensive medication

  2. (B)

    improvement: reduction but not normalization of BP centile (> 95th centile) while on drug therapy, with or without previous intervention

  3. (C)

    unchanged: same BP centile with the same or increased drug therapy, with or without previous intervention

  4. (D)

    failure: technical failure of intervention [23].

The renal function was assessed as estimated glomerular filtration rate (eGFR) with the Schwartz formula [24]. Renal dysfunction was present in patients with eGFR < 90 ml/min per 1.73 m2 body surface area.

Results

Of 36 patients with MAS, 64% were male and 77% were Caucasian. Age at presentation with HTN varied from 10 days to 10 years, with a median age of 2.7 years. The median age of referral to our centre was 4 years (range 1 month–13 years). The median time from onset of symptoms to referral to our unit was nearly 3 months (range 3 days–5.7 years).

Presenting features

BP at presentation was above the 99th centile for age, gender and height centile in all children [median systolic BP of 160 (range 117–250) mmHg]. Hypertension was incidentally found in 25% (nine) of patients at a routine medical examination or during investigations for other pathological conditions, whereas another 25% (nine) presented with symptoms of congestive cardiac failure and 17% (6) with hypertensive encephalopathy. Hypertension was detected in eight children (22%) at screening, due to known or suspected genetic syndromes. Other modes of presentation included claudication and failure to thrive. No patients had a history suggesting chronic inflammation and/or elevated levels of inflammatory markers. Physical signs of vascular abnormality could be detected in 50% (18) of patients: 14% (five) had a heart murmur, 19% (seven) an abdominal bruit, 6% (two) children demonstrated decreased or absent femoral pulses, and 11% (four) of patients presented with two or three of the above signs. Because of the difficulty of obtaining BP in the legs, especially for small children, the differential in the BP between the upper and the lower extremities was not routinely achieved.

Left ventricular hypertrophy was diagnosed in 80% (26/32) of children, and hypertensive retinopathy in 21% (5/24). Renal involvement at onset was investigated in all patients, with renal dysfunction (eGFR below 90 ml/min per 1.73 m2) found in 17% (six) of children: median eGFR 103 (range 53–161) ml/min per 1.73 m2. Technetium-99 m-dimercaptosuccinic acid (DMSA) scintigraphy detected pathological differential up-take of radionuclide below 45% in one kidney in 59% of patients.

Associated findings

Genetic conditions were present in 36% (13) of patients. Neurofibromatosis type I (NF-1) was the most common associated genetic syndrome (19%). Of the children with NF-1, 8% (three) were also affected by neoplastic disease (Table 1). The presence of a neoplasm was demonstrated in one patient with Feuerstein–Mims syndrome and in 8% (three) of children without a defined genetic condition. In these cases a large abdominal malignant tumour was the cause of the vascular disease, with compression of the aorta and its visceral branches. In all cases the mass had been removed, but the children remained severely hypertensive.

Table 1 Associated findings
Full size table

Scoliosis was detected in 14% (five) of patients, and altered skin pigmentation (hyperpigmentation or hypopigmentation) in 22% (eight). Neurodevelopmental delay, convulsive episodes, impaired visual or hearing function were present in 28% (ten) of patients, related to their underlying diagnosis and/or cerebrovascular disease, and/or degree of hypertension.

Extent of disease

All patients had narrowing of the abdominal aorta at various segments, associated with other vascular abnormalities. Renal artery stenosis (RAS) was present in 86% (31) of children, and bilateral RAS in 50% (18). Intrarenal vascular involvement was detected in 31% (11) of patients, all of whom also had main RAS, which was bilateral in eight cases.

Intestinal arterial involvement was found in 67% (24) of children. There was a simultaneous narrowing of two or three branches in 19% (seven) of cases and of the iliac arteries in 8% (three). However, none of them was symptomatic of mesenteric ischaemia. Cerebrovascular disease was diagnosed by cerebral perfusion studies: technetium (Tc)-99 m HMPAO (single-photon emission computed tomography with hexamethyl-propylene amine oxide-labelled with technetium-99), cerebral computed tomography and magnetic resonance angiography. Of the patients, 44% (16) demonstrated abnormal head or neck arteries, with narrowing of either the carotid or subclavian artery in 14% (five) of children, one or several abnormal intracranial arteries in 31% (11), and 11% (four) with cerebral infarctions.

Outcome

Over a median follow-up period of 4.5 (range 1–19.7) years, patients underwent multiple therapeutic interventions. Anti-hypertensive medications represented the first-line treatment for all patients, with a median number of drugs of 4 (range 2–7).

Medical therapy

Six children received medical treatment only (16%); three patients did not need further treatment because good BP control was achieved with anti-hypertensive agents alone (Table 2). In two children the high complexity and widespread nature of the arterial disease excluded the possibility of a curative angioplasty or surgical intervention, and, in one case, the study of the arterial tree was not clear enough to allow invasive treatment (renal arteries not identified on angiography performed in 1991).

Table 2 Long-term outcome. BP response to treatment was evaluated as (A) cure—normalization of BP; (B) improvement—reduction in BP centile while on drug therapy, with or without previous intervention; (C) unchanged; (D) failure—technical failure of intervention (PTA percutaneous transluminal angioplasty)
Full size table

The three children who had initially responded well to drug therapy, obtained good BP control and required fewer anti-hypertensive agents. However, the other three children were felt not to be amenable to angioplasty or surgery and had had a more severe outcome. One child with widespread vascular disease died from a cerebrovascular accident (CVA), another developed claudication and worsening renovascular disease. The BP of the third patient was less than the 95th centile, with evidence of decreased perfusion on a single-photon emission computed tomography (SPECT) cerebral scan.

Percutaneous transluminal angioplasty

Of the patients, 36% (13) underwent percutaneous transluminal angioplasty (PTA) at a median age of 8.6 (range 0.3–15.6) years (Table 2). PTA was performed more than once in six patients, with a mean of 1.6 interventions. Five patients underwent PTA for unilateral or bilateral RAS with stenting in all but one patient. Five children had dilatation of the aorta, which was also stented in three of them. PTA of both renal arteries and aorta (in two cases also stented) was performed in three children.

Nine patients achieved better BP control, and one child was normotensive without anti-hypertensive medications. Three patients did not obtain any improvement of their HTN: one patient where no distal renal arteries could be detected on renal angiography, one patient with a re-stenosis of the aorta after angioplasty, and one child who had undergone two attempts of PTA that did not lead to improved BP. For this patient, the parents refused surgical intervention.

PTA plus surgical treatment

Among patients treated with PTA for the narrowing of their arteries, seven subsequently also needed surgical treatment (Table 2). The median age at PTA was 6.5 (0.8–10.1) years, whereas the median age for the surgical intervention was 7.6 (1.8–10.3) years. For all patients, the reason for further surgery was failure of angioplasty to improve BP control and, in three children, also worsening renal function.

PTA for either unilateral or bilateral RAS was performed in three children, with stenting in one case. Three patients underwent balloon dilatation of the aorta. In one case the hypertension was treated through dilatation and stenting of both renal arteries and aorta.

Surgical procedures included autologous surgery with renal revascularization (one), synthetic graft interposition for renal revascularization (two) and aortic reconstruction (two) and nephrectomy (two). Owing to persistent severe HTN, a following PTA was necessary for three patients.

At the end of the follow-up period, 28% (2/7) of patients were normotensive without the use of anti-hypertensive agents; two children had improved BP on anti-hypertensive medications. In two patients HTN continued, and one child died from complications of the intervention.

Surgical treatment

In 28% (ten) of patients surgery was the only therapeutic approach for the treatment of MAS, after a first attempt with medical treatment (Table 2). The median age at surgery was 4.7 (0.9–10.5) years. The indication for surgery was severe HTN (systolic BP up to 300 mmHg) which was associated with renal dysfunction in five patients. In one case the high complexity of vascular disease was the main reason for surgery; in another three children the surgery represented the only possible approach (therapeutic strategy performed more than 20 years ago). The surgical interventions included autologous surgery for aortic reconstruction (one) and renal revascularization (three), synthetic graft interposition for renal revascularization (one) and aortic reconstruction (one), and nephrectomy (four). Only one patient was treated with angioplasty after the surgical intervention, and only one needed a second surgical procedure.

At the end of the follow-up period, six patients were off anti-hypertension therapy, five with good BP control (median BP centile: 65th, range 50th–90th ) and one with gradually increasing BP due to a new vascular stenosis, needing a further intervention.

Long-term outcome of all patients

The mortality rate was 8% (3/36 patients). One patient with severe bilateral RAS, intrarenal disease and complex vasculopathy managed in the early 1980s was judged not to be amenable to surgery and died after a CVA. The second patient was diagnosed with severe HTN (systolic BP 190 mmHg) at the age of 3 years and cerebrovascular disease and died after two CVAs before the planned surgery could take place. The third death was a patient with bilateral RAS, intrarenal disease and cerebrovascular involvement.

At the end of the follow-up time, 25% (nine) of patients were off therapy (A), 58% (21) still required anti-hypertensive therapy with better BP control (B), 14% (five) of patients continued to have HTN, despite treatment (C), 3% (one) was hypertensive for a technically unsuccessful intervention (D) (Table 2).

Of 33 patients who had survived, 90% (30) obtained an improvement in their HTN, 27 with normal renal function. A total of eight patients (24%) had renal dysfunction at the end of the study (eGFR median value 71 ml/min per 1.73 m2 body surface area, range 52–87 ml/min per 1.73 m2). There was no significant difference between baseline and final eGFR (P > 0.1).

In the three patients with persistent HTN, two had bilateral RAS and intrarenal disease. Two children had cerebrovascular involvement, one child had chronic kidney disease (CKD; eGFR 55 ml/min per 1.73 m2). All these children underwent multiple invasive treatments during the study period.

Discussion

This series of 36 children affected by MAS is the largest series to date. All children also suffered from abnormalities in other vascular beds, including 86% with RAS. The vasculature that was felt to be most important for the HTN was treated, and a successful outcome (defined as cured or improved BP) was demonstrated in 90% of patients, although 24% of patients had CKD at the end of a median follow-up time of 4.5 years.

Several studies have documented the clinical features of MAS, thereby contributing to our knowledge of this condition (Table 3). Some of them referred only to idiopathic MAS [1, 25, 26], while other studies included cases of MAS of different aetiologies [10, 27, 28].

Table 3 Review of the literature
Full size table

A new finding in our study was the low median age at presentation with HTN: 2.7 years, with 33% (12) of patients being diagnosed before they were 8 months old. Previous studies have reported MAS as a disease of late childhood or early adulthood [1, 29], but, more recently, a lower age at diagnosis has been reported, with a mean age of 4.5 years [25]. The trend towards diagnosis in younger children is likely to be the result of more severe cases from tertiary and quaternary referrals and improved diagnosis for earlier detection of this condition. It is also likely that children previously died before diagnosis and/or intervention.

In this series, MAS often presented with severe symptoms, although it was also diagnosed as an incidental finding in 25% of patients. Many children had involvement of the renal arteries as well as of other major visceral branches, although clinically evident mesenteric ischaemia was not detected in any child, presumably due to a rich collateral circulation. Almost 45% of our children had cerebrovascular disease [11–14], and, in two of the three children who died, the cause of death was due to CVA. Other studies have not reported such a widespread vascular involvement. Visceral arteries have almost invariably been shown to be abnormal [1, 10, 26, 27], but the presence of cerebrovascular disease was rarely recognized in previously reported children [30]. Our data show that it is important to look for cerebrovascular abnormalities in children affected by MAS.

Previous studies emphasized surgical therapy as the best approach for children affected by MAS [14], with early normalization of BP after open surgery in 94% of all cases [31]. This is similar to that for our single-centre patient cohort [32], where 86% of patients had surgery in conjunction with conservative therapy. The results of our study appear to confirm better outcomes for those patients whose MAS was felt to be amenable to surgery alone without PTA. However, this is only a case series, so there may be a selection bias between different treatment groups and changes over time with improvements in angioplasty techniques over the years.

In our study, PTA was a very successful treatment in a proportion of affected children, while, in most previous studies, it has failed to provide a lasting clinical and angiographic improvement [10, 27, 28]. In three children with MAS, PTA with stenting was proven to be effective, in the short to medium term (less then 1 year), by D’Souza et al. [33]. They suggested a possible therapeutic approach to obtain a prompt reduction of extremely high BP, in selected cases not involving renal and mesenteric arteries [34]. In a recent retrospective review at our centre [35], PTA was successful in reducing BP in 55% (18/33) of patients, using different technical approaches for angioplasty of renal arteries and/or aorta.

In our study described now, RVH was treated with PTA of renal arteries alone in eight children, aortoplasty in another eight, and with combined PTA for both renal arteries and abdominal aorta in four children. This approach was decided on the basis of which blood vessel was judged to be most important for causing HTN.

The possibility of medical therapy was reported by Panayiotopoulos et al. [10], who describe an initial moderate to good BP response in patients with MAS. Other authors claim that medical management of MAS with moderate arterial impairment can allow BP control and reduce the risk for complications [6, 7, 25, 34, 36]. In our study, medical therapy had an important role in the management of MAS. There was a small group of children with not too complex and widespread a vasculopathy where well-controlled BP seemed to be achieved with anti-hypertensive agents alone.

We propose a different stepwise treatment policy of MAS, which does not emphasize a single therapeutic option (PTA or surgery) as the best choices for these children, but includes all these treatments in a structured therapeutic pathway based on the clinical characteristics of patients and validated by a long-lasting multi-disciplinary approach to this condition. Figure 1 suggests a therapeutic algorithm based on our current approach to children with MAS.

Fig. 1
figure 1

Therapeutic algorithm

Full size image

Our clinical preference is to delay surgical revascularization until the child is as developed as the clinical situation will allow, preferably to the stage of adolescence. This is because prosthetic conduits are preferable to the saphenous vein, which has a small but real risk of late aneurysmal degeneration. Obviously, prosthetic conduits have no growth potential, and their use in smaller younger children may require a further secondary surgical reconstruction at or after adolescence. Renal stent angioplasty, in our experience, carries a significant in-stent stenosis rate of up to 50%, but it has not compromised the potential for surgical reconstruction in any of the children of our experience. Our policy is, therefore, to use stents, even in younger children, accepting that many will need later re-intervention and eventual surgery but in a bigger child where prosthetic renal revascularization stands a better chance of being the definitive procedure for that individual’s life span.

Conclusions

We report a group of children with MAS that are younger and with more widespread disease then previously described in the literature. Individualized use of medical, interventional and surgical treatments managed to control or improve the BP in 90% of the children.