Longitudinal Evaluation of Segmental Arterial Mediolysis in Splanchnic Arteries: Case Series and Systematic Review

Background Segmental arterial mediolysis (SAM) is a rare non-atherosclerotic, non-inflammatory vascular disorder varying widely in clinical course. The purpose of this study is to analyze detailing clinical and imaging manifestations over time in patients with SAM through a literature review and to suggest an optimal management strategy. Methods A retrospective review of eight consecutive patients diagnosed with SAM between January, 2000 and January, 2012 was conducted. All presented with acute-onset abdominal or flank pain. Clinical features, imaging studies, and laboratory findings served as grounds for diagnosis, having excluded more common conditions (ie, fibromuscular dysplasia, collagen vascular disorders, or arteritis). CT angiography was done initially and repeated periodically (Week 1, Month 3, then yearly). Treatment was conservative, utilizing endovascular intervention as warranted by CT diagnostics. In a related systematic review, all English literature from 1976 to 2015 was screened via the PubMed database, assessing patient demographics, affected arteries, clinical presentations, and treatment methods. Findings Ultimately, 25 arterial lesions identified in eight patients (median age, 62.8 years; range, 40–84 years) were monitored for a median period of 26 months (range, 15–57 months). At baseline, celiac axis (3/8, 37.5%), superior mesenteric (4/8, 50%), and common hepatic (2/8, 25%) arteries were involved, in addition to isolated lesions of right renal, splenic, right colic, middle colic, gastroduodenal, left gastric, right gastroepiploic, proper hepatic, right hepatic, and left hepatic arteries. Compared with prior publications, celiac axis and superior mesenteric artery were more commonly affected in cohort. Arterial dissections (n = 8), aneurysms (n = 5), stenoses or occlusions (n = 4), and a single pseudoaneurysm were documented. Despite careful conservative management, new splanchnic arterial lesions (n = 4) arose during follow-up. Considering the few available reports of new arterial lesions in the literature, newly developing pathology is a distinctive feature of our patients, four of whom eventually required endovascular interventions. Conclusions Careful clinical observation via periodic CT angiography is required in patients with SAM, checking for newly developing lesions. The natural history of SAM should be clarified in a larger patient population.


Introduction
Segmental arterial mediolysis (SAM) is a rare non-atherosclerotic, non-inflammatory vascular disorder that varies considerably in clinical course. Its`etiology remains unclear, although some consider it a variant or a precursor of fibromuscular dysplasia (FMD) [1,2]. Others have suggested that SAM is a secondary phenomenon, resulting from vasospasm or arterial wall injury due to immune complexes [3]. Originally described in 1976 by Slavin et al [4], only around 100 cases of SAM have since been reported in the literature [5].
SAM is typically manifested in splanchnic arteries where lytic degeneration of medial smooth muscle occurs, culminating in tearing/separation from adventitia and adjacent fibrosis. Patients variably present with dissection, aneurysm, stenosis, occlusion, or hemorrhage after rupture, often calling for emergency surgical or endovascular intervention. Unfortunately, no standard criteria exist at present for differentiating SAM from inflammatory vasculitis, and the clinical course of SAM is varie, with no clear tendency to progress, resolve, or stabilize. Likewise, there are no established therapeutic or monitoring guidelines stipulating circumstances where surgical or other interventions are indicated.
The purpose of this study is to review our experience with SAM management and to suggest an optimal management and surveillance strategy in SAM through thorough literature review.

Methods
A retrospective review of eight consecutive patients (male, 4; female, 4) diagnosed with SAM between January, 2000 and January, 2012 was conducted with approval of the Institutional Review Board (IRB No SNUH-1408-027-601). The obtainment of informed consent was waived and patient information/records were anonymized and de-identified prior to analysis. anticoagulants or antiplatelet agents. CT angiographic studies were repeated at Week 1 and at Month 3 after diagnosis and thereafter were done yearly. On occasion, CT angiography was ordered more often (at physician's discretion) if new lesions or symptoms developed. In patients requiring superior mesenteric arterial stenting, self-expandable stents were placed in a manner reported previously [7].

Literature review
PubMed databases were searched for all articles published in English between January 1, 1976 and August 31, 2015. "Segmental arterial mediolysis" and "segmental mediolytic arteritis" served as keyword combinations. All abstracts, case reports, patient series and citations scanned were examined. Extraction of study data was achived using a standardized template, which included author`s name, publication year, patient demographics, affected arteries, clinical presentations, and treatment methods. A full description of search terms, strategy and screening stages are provided in S1 Table. This literature review follows the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (S1 PRISMA Checklist, S1 PRISMA Flow Diagram).

Statistical analysis
We analyzed variables of our case series and of those in the literature. Because events were relatively few and biased, the binomial Clopper Pearson exact method was used to generate 95% confidence intervals.
During follow-up, all patients underwent CT angiography periodically in accord with study protocol. Upon conclusion, 25 lesions ultimately accrued, including three additional aneurysms (SMA branch, pancreaticoduodenal and ilieocolic arteries), marginal artery occlusion, and bilateral renal arterial dissections (Table 2).

Surveillance results
Although patient No. 8 underwent endovascular embolization of an initially identified gastroduodenal pseudoaneurysm (Fig 1), other patients opted for regular monitoring. At completion of the follow-up period, lesions in four patients were stable, whereas new arterial lesions developed in the other four (No. 1, 2, 5, 6). In one patient (No. 1), a SMA branch aneurysm and a marginal arterial occlusion appeared at Month 2. Another patient (No. 2), presenting with SMA dissection, displayed dissections of right and left renal arteries with ectasia at Week 1 ( Fig  2). Patient No. 5, also presented with SMA dissection and developed a pancreaticoduodenal aneurysm by Day 9. Finally, patient No. 6 developed an ileocolic aneurysm at Month 1 (Fig 3). Among these patients, symptomatic isolated dissection of SMA (SIDSMA) was first considered in three (No. 1, 2, 5). However, a diagnosis of SAM was made after other splanchnic arterial lesions materialized. Two patients (No. 1, 5) required stent placement due to progression of dissection, as specified in our previously published recommendations [7].

Literature cases
To date there are 76 studies and 101 cases of SAM (Tables 3 and 4), of which 60 were men and 41 were women. Median age was 56.9 years (range, 0-91 years 95% CI range,). Most patients presented with abdominal pain (68%), whereas eight patients were asymptomatic. In 68 patients (67%), diagnosis was confirmed by histologic means. Others were diagnosed clinically. The vessel most often associated with SAM was splenic artery, accounting for 28% of all SAM involvement. Common hepatic artery (and hepatic branches), celiac trunk, and renal arteries were the next most commonly involved.
Of the 94 cases with reported survival outcomes, SAM-related mortality was 22% (21 patients), with 78% (73 patients) surviving acute presentations of SAM. Of the 21 patients who died of SAM-related causes, 13 reportedly died before any intervention could be attempted (i.e., death on arrival to hospital) or without intervention, and five died despite attempts at nonoperative conservative management.
Eventually only 14 patients were conservatively managed. Another 47 patients underwent surgical ligation of bleeding vessels/or resection of aneurysmal segment, often with vascular reconstruction either as primary management or after failed endovascular intervention. Twenty-four patients were managed by endovascular intervention through sole use of coil embolization.
In most cases reviewed, follow-up indicated that patients were largely asymptomatic clinically, with additional imaging showing variable outcomes from complete resolution to no change in untreated aneurysm [5]. Few reports of new arterial lesions appears in the literature. Slavin et al (1976) were the first to define SAM, describing this pathologically distinct entity based on three autopsies. Fundamentally, SAM involves lytic degeneration of medial smooth muscle leading to tearing/separation from adventitia and adjacent fibrosis. Hence, patients commonly present with aneurysms, dissections, stenoses and ruptures, often requiring emergency surgical or endovascular intervention [32].

Discussion
The 101 cases of SAM that have been reported thus far likely represent a gross underestimate of its true incidence. Typically, splanchnic vessels of middle-aged and elderly adults are affected in SAM, although carotid, renal, intracranial, and iliac arterial involvement has been reported, and some cases have developed in young individuals. To date, some authorities insist that SAM represents a variant or a precursor of FMD [1, 2]. However, young females are usually afflicted with FMD, showing diffuse disarray of media in mid-and distal arterial segments where smooth muscle is replaced by collagen [6,83]. Arterial stenoses are common in FMD, but aneurysms and dissections are rare [35,83]. The differential diagnosis also includes Marfan's syndrome, polyarteritis nodosa, Takayasu's arteritis, Behçet's disease, allergic granulomatous angiitis, and various disorders of collagen (eg, Ehlers-Danlos syndrome, Loeys-Dietz syndrome, and neurofibromatosis) [35]. SAM differs significantly in terms of pathologic findings, laboratory abnormalities, and sites of involvement, predominantly involving splanchnic vasculature and presenting distinctively. Marfan's and Ehlers-Danlos syndromes commonly manifest as dissections and aneurysms. Loeys-Dietz syndrome also is usually marked by aneurysms. Although neurofibromatosis may produce long-segment stenosis and aneurysms, dissections are rare [35].
In a literature review, abdominal pain was the most common presenting symptom (68%), followed by hemodynamic shock (25%), neurologic symptom (12%), and 11 patients (11%) died before further investigation and/or attempted management. However, eight cases (8%) reported an asymptomatic presentation. In our review, seven of our patients (87.5%) presented with acute abdominal pain and one patient experienced acute flank pain.
Mean age at presentation was slightly higher in our patients, compared with the pooled data of literature review (62.8 vs 56.9 years). Multiple vessels were initially involved in five of our patients (62.5%), as opposed to 47% in other reports [54,84]. Some sources further suggest that multiple splanchnic vessels are affected during the course of SAM [32]. In support of this tenet, half of our patients developed new lesions of splanchnic arteries (Pt. No. 1, 2, 5, 6) (Table 1). Consequently, careful clinical follow-up is recommended, checking regularly for new arterial lesions via CT angiography. According to our protocol, CT angiography is done at presentation and repeated periodically (Week 1 and Month 3 after diagnosis, then yearly). Only if Three of our patients (37.5%) were initially considered as SIDSMA in our series. Once new splanchnic arterial lesions materialized, a diagnosis of SAM was eventually established. It may well be that SIDSMA is a preliminary manifestation of SAM. Regular imaging and careful monitoring are thus prudent in instances of SIDSMA as well.
No formal guidelines for management of SAM exist as yet. Patients with shock and intraabdominal hemorrhage should be treated with emergency surgical or endovascular intervention. If a lesion progresses, increasing the risk of organ ischemia, endovascular intervention is perhaps the foremost therapeutic option [32]. Previous studies have indicated that a benign course of the disease requires no therapy [52]. Given that the arterial walls are already prone to dissection or development of aneurysms, intra-arterial catheter manipulation and balloon dilatation stand to worsen or instigate arterial dissections. Accordingly, invasive procedures may be reserved for patients with hemodynamically unstable conditions or significant end-organ ischemia [47,63,69,85]. Four of our patients submitted to endovascular interventions. Two had stents placed in SMA, adhering to related treatment guidelines of our group; [7] one underwent coil embolization for a pseudoaneurysm of gastroduodenal artery with bleeding; and in another we opted for right renal arterial stent insertion due to progressing azotemia. No surgical interventions were done. The utility of corticosteroids in the management of this disease is questionable, given the absence of inflammation in histologic preparations. Active management of hypertension may otherwise be beneficial [62].
The present study has several limitations. As other case series reports, our initiative deals with a small group of patients (N = 8). The latter may seem abundant by comparison, relative to 101 similar case reports in the literature, but small patient samplings provide only limited perspective of this disease entity. In addition, none of the arterial lesions had pathologic confirmation, which would have been prohibitive. Each diagnosis was based solely on clinical and imaging findings, in conjunction with an absence of laboratory abnormalities and exclusion of diagnostic alternatives [6]. We applied diagnostic criteria of Kalva et al for this study. Nevertheless, these criteria are in need of further validation. The present study was also retrospective in nature, relying upon only eight patients. However, in the only prior study of more than eight patients, CT angiographic data were insufficient. By comparison, our efforts have generated a significant body of information on SAM, addressing diagnosis, therapy, and follow-up management. With increasing awareness of SAM, new studies are coming forward. The clear

Author Contributions
Conceived and designed the experiments: SM SKM.
Performed the experiments: HSK SM.
Analyzed the data: SM AH CC JH.
Contributed reagents/materials/analysis tools: HSK SM AH CC SKM JH.
Wrote the paper: HSK SM.