Introduction

The incidence of true splenic artery aneurysms (SAAs) has been estimated at 0.5-2%, representing approximately 50-60% of all visceral artery aneurysms (VAAs). SAA rupture is associated with a 10-25% mortality rate [1-3]. Endovascular procedures have become the preferred method for treating SAAs due to their high efficacy and low rate of complications [4]. Endovascular procedures can be divided into 2 types: those that occlude the entire artery and those that only occlude the aneurysm sac while preserving the parent artery. Although the latter method has a lower complication rate, it can still result in splenic ischaemia, post-embolisation syndrome, splenic abscess, or portal vein thrombosis [5] and requires periodic follow-up imaging to assess possible reperfusion, compaction of embolic material, and enlargement of the aneurysmal sac [6,7]. Different endovascular devices suit both techniques, but coil embolisation, either simple or stent-assisted, is the method most frequently used.

According to the latest Society for Vascular Surgery (SVS) guidelines from 2020, follow-up examinations in patients with true SAAs treated with coil embolisation should be performed with the use of computed tomography angiography (CTA), ultrasound, or magnetic resonance angiography (MRA) [8]. However, this recommendation is weak (strength of recommendation: 2) and based on moderate-quality evidence (quality of evidence: B). We aimed to systematically evaluate the utility of different imaging modalities: digital subtraction angiography (DSA), CTA, MRA, contrast-enhanced ultrasound (CEUS), and duplex ultrasound (DUS) for follow-up screening of patients with SAAs treated with coil embolisation.

Material and methods

This systematic review was conducted according to the PRISMA 2020 Statement [9].

We included primary original clinical research studies written in English, published after 2005, which assessed various imaging techniques as follow-up modalities after coil embolisation of true SAAs, with a minimum sample size of 10 patients. Systematic reviews, meta-analyses, narrative reviews, case reports, letters to editors, commentaries, conference abstracts, guidelines/statements, expert opinions, preprints, and book chapters were excluded. Five databases were included in the search: Embase, Medline Ultimate, PubMed, Scopus, and Web of Science, with the following query: “((DSA) OR (digital subtraction angiography) OR (angiography) OR (MRA) OR (magnetic resonance angiography) OR (CEMRI) OR (CE-MRI) OR (MRI) OR (magnetic resonance) OR (CTA) OR (computed tomography angiography) OR (CECT) OR (CE-CT) OR (CECTA) OR (CE-CTA) OR (CT) OR (computed tomography) OR (CEUS) OR (CE-US) OR (contrastenhanced ultrasound) OR (ultrasound)) AND ((follow-up modality) OR (follow up modality) OR (follow-up) OR (follow up) OR (follow-up imaging) OR (follow up imaging)) AND ((embolization) OR (coil embolization) OR (coiling) OR (selective coil) OR (selective coil embolization) OR (intravascular treatment)) AND ((splenic artery aneurysm) OR (SAA) OR (visceral artery aneurysm) OR (splanchnic artery aneurysm) OR (splenic artery aneurism) OR (visceral artery aneurism) OR (splanchnic artery aneurism))”, which yielded a total of 1704 records. All databases were last searched on 10 April 2024. At each screening stage, 2 independent screeners reviewed the results. Final decisions in uncertain cases were reached by consensus between the 2 screeners. Data from each included work were extracted by one extractor.

The following data were extracted from each included study: study characteristics (sample sizes, methodology used, year of publication) and findings related to the efficacy of a specific imaging modality.

After removing 874 duplicates, the remaining 866 records were screened by title and abstract. This resulted in 43 records that met the predefined inclusion and exclusion criteria, and the full-text versions of all these studies were obtained. Entire data reports were evaluated for eligibility, from which 23 studies were excluded due to (i) having only an abstract without a full-length article (n = 9), (ii) inappropriateness (n = 2), (iii) language other than English (n = 2), and (iv) sample size too small (n = 10). Eventually, 20 original and relevant studies were included in this review (Figure 1).

Figure 1

Flowchart for the selection process

We divided the studies into the following parts: (i) studies using DSA as one of the follow-up imaging modalities, (ii) studies using MRA as one of the follow-up imaging modalities and not discussed before, (iii) studies using DUS as one of the follow-up imaging modalities and not discussed before, and (iv) studies using CTA as the only follow-up imaging modality.

Results

Discussion

The latest Cardiovascular and Interventional Radiological Society of Europe (CIRSE) standard of practice recommends imaging examinations at specific intervals following endovascular treatment – namely, at one month, 12 months, and annually after that [29]. However, the literature lacks evidence on the sensitivity and specificity of the most commonly used imaging modalities in this context [30].

DSA is the most established and oldest follow-up modality for patients undergoing endovascular treatment for SAAs. However, it is an invasive procedure that involves ionising radiation, and small instances of aneurysmal sac reperfusion can be obscured by radio-opaque coils. DSA has been extensively studied, with numerous investigations examining the relationship between packing density, coil compaction, and sac reperfusion. This method has also been widely employed in neuro-interventions to monitor the treatment of intracranial aneurysms, from which the Roy et al. classification has been adapted in several of the reviewed studies to evaluate visceral artery aneurysm reperfusion and coil compaction following treatment.

For example, in the context of splenic artery aneurysms, Yasumoto et al. [11] demonstrated that a packing density exceeding 24% is sufficient to prevent reperfusion during long-term follow-up. At the same time, Wojtaszek et al. [13] reported that a packing density of over 29% is necessary when using detachable coils for these aneurysms.

Our review evaluated 20 non-randomised, retrospective studies concerning the endovascular treatment of SAAs. In 4 of these studies, the authors employed a single imaging modality (CTA: n = 3, MRA: n = 1) to assess treatment outcomes; in the remaining studies, multiple modalities were utilised. Specifically, CTA was employed in 17 studies, MRA in 13, DSA in 5, and DUS in 12. Unfortunately, many publications failed to specify the imaging methods used for individual cases, complicating comparisons, particularly regarding the assessment of sac reperfusion, its potential severity, its clinical impact, and the sensitivities and specificities of each imaging technique.

Nonetheless, 4 studies provided more comprehensive data on follow-up modalities. Wojtaszek et al. [13] and Lamparski et al. [14], like Iryo et al. [31], observed that MRA is more effective than DSA in detecting small aneurysmal sac reperfusion. Furthermore, Lamparski et al. [14] evaluated the utility of DUS, which demonstrated high specificity for detecting type III aneurysm sac reperfusion as classified by Roy et al.; however, the method’s sensitivity was regarded as relatively low. Kawai et al. [12] reported that all performed MRA studies were diagnostic, with a 100% agreement rate when compared to DSA. In the same study, CTA was conducted in 11 out of 21 cases; however, in 9 instances, CTA was deemed unevaluable due to coil artifacts. Among the remaining cases, one resulted in a true positive and one in a false positive result.

CTA was utilised as a follow-up modality in an additional 16 studies, employed as a standalone method in 2 instances, while the remainder were combined with other imaging techniques. In 11 of these studies, the authors noted challenges in evaluating treatment efficacy due to beam hardening artifacts. Tulsyan et al. [15] categorised the severity of artifacts in CTA examinations into 3 groups: grade 1 for minor, grade 2 for moderate, and grade 3 for severe artifacts. In alignment with findings from other studies, most CTA scans (17 out of 33) exhibited severe artifacts, with 11 cases demonstrating moderate artifacts and only 5 cases showing minor artifacts.

Despite its limitations, CTA continues to be a valuable diagnostic tool for patients immediately following the procedure, particularly when severe complications such as bleeding or organ ischaemia are suspected [32,33]. However, the significant beam hardening artifacts produced by endovascular devices – primarily coils and tantalum-based liquid embolic agents – have led several authors to question the reliability of CTA in accurately assessing persistent reperfusion of the aneurysmal sac or coil compaction [34].

Unlike the previously mentioned methods, MRA is a promising non-invasive technique that does not involve ionising radiation. Several studies have demonstrated the superiority of MRA over DSA in detecting minute aneurysmal sac reperfusion. Iryo et al. [31] reported that MRA performed after the embolisation procedure had 93% agreement with DSA. In comparison, Wojtaszek et al. [13] showed a 33% superiority of MRA over DSA in 9 patients needing reintervention after SAAs coil embolisation. Like DSA and CTA, MRA imaging enables the evaluation of collateral circulation, which is crucial for monitoring patients post-treatment, because increased blood flow through these vessels can elevate arterial wall stress and potentially result in the development of secondary aneurysms [35].

While CTA and MRA have been partially evaluated as alternatives to DSA, evidence regarding the utility of DUS in the follow-up of patients undergoing endovascular treatment of SAAs remains limited. DUS is a low-cost, non-invasive, and safe technique; however, it is constrained by limitations such as operator dependence and reduced effectiveness in patients with obesity or intestinal gas, which can impair repeatability and diagnostic accuracy. Lamparski et al. [13] reported a positive predictive value of 75.0% for DUS in identifying cases that required re-embolisation, suggesting its potential as a follow-up tool for monitoring selected low-risk patients after selective embolisation of SAAs, especially in those contraindicated for other imaging modalities. Adding CEUS to standard DUS presents a promising solution. However, to our knowledge, there is a lack of evidence demonstrating the superiority of CEUS over conventional DUS for post-endovascular follow-up in patients with SAAs.

Conclusions

Endovascular treatment using coils in patients with SAAs is highly effective and has a relatively low complication rate. The optimal method of elective follow-up should be widely available, noninvasive, reproducible, and accurate. Although DSA has relatively high sensitivity and specificity in detecting recanalisation after endovascular treatment, it requires radiation and is invasive. Therefore, it should not be performed as a routine check-up.

Several studies have proven that MRA is superior to DSA in detecting aneurysmal sac reperfusion, particularly small residual neck recanalisation. Given the above information and that MRA is a noninvasive modality, we firmly believe it should be used as a method of choice in patients with SAAs undergoing coil embolisation.

Although considered rare, serious complications such as abdominal bleeding or significant organ ischaemia can occur after coil embolisation. Due to its availability, short examination time, and effectiveness in detecting abdominal bleeding, CTA is undoubtedly the method of choice in such cases. However, due to excessive beam hardening artifacts generated by endovascular devices, it should not be considered a modality of elective follow-up after treatment. Furthermore, CTA is contraindicated in pregnant women.

Due to its low sensitivity and specificity, DUS cannot be used as a single modality; however, in specific clinical scenarios, it may reduce the number of MRA images performed after the endovascular treatment.

Altogether, the available evidence regarding the follow-up imaging methods after SAAs coil embolisation is limited. The recommendations are based on low-quality research, which does not directly address the issue of the best possible follow-up modality, as demonstrated in Tables 1-4, where we added the information about the “level of evidence” of a particular study, according to the Oxford Centre for Evidence-Based Medicine: Levels of Evidence (March 2009). To conclude, there is a need for high-quality research with larger sample sizes to unambiguously indicate the imaging modality of choice for follow-up in patients with SAAs after coil embolisation.