Pharmacotherapeutics for prevention of aortic root enlargement in Marfan Syndrome – A network meta-analysis of randomized controlled trials

Marfan syndrome (MFS) is a life-threatening autosomal dominant condition with a prevalence ranging from one every 3000 to one every 5000 persons.¹ Aortic root dissection due to progressive aortic enlargement is the leading cause of premature death in MFS. There is no curative therapy for MFS, and aortic dilatation is the primary therapeutic target in these patients. Despite this, the lifespan of MFS patients has dramatically increased from 32 to 72 years, and this has mostly been attributed to early diagnosis, improved lifestyle guidance, and non-invasive aortic imaging along with elective aortic root repair.^2,3 The evidence for the medical treatment in reducing aortic root dilatation remains unclear. Beta-blockers are considered first-line for the prevention of aortic root dilatation despite the inconsistent evidence. More recently, angiotensin receptor blockers (ARBs) have been tested with limited success in several trials due to their antagonizing properties against transforming growth factor-beta (TGF-b).^4,–6 Habashi et al. in 2006 found that the clinical manifestations of MFS are due to altered activation and signaling of TGF-b and losartan was noted to antagonize the TGF-b signaling in a mouse model.⁷ In MFS, a defect in FBN1 glycoprotein leads to a reduction or alteration in fibrillin-1, which can cause an increase in the level and activity of TGF-b. Several studies failed to show an advantage of beta-blockers or ARBs over each other for the prevention of aortic root growth (ARG), adding to the confusion as to which is the preferred treatment choice. Therefore, the objective of this network meta-analysis from all the randomized controlled trials (RCTs) is to synthesize evidence from existing studies and to compare the efficacy of various therapeutic options for the prevention of aortic root dilatation in patients with MFS.

We performed a comprehensive literature search in PubMed, Cochrane library, and Embase. The keywords utilized included ‘Marfan syndrome’ with the limits for human and RCTs. The RCTs were considered for inclusion in our meta-analysis if they studied the effects of medical therapy for the prevention of aortic root dilatation. Our outcome of interest was the change in aortic root dilatation. Throughout this process, we followed Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines for all stages of the design and implementation. We extracted the relevant data from the finally selected studies.

We used the overall change in aortic root dilatation as a continuous variable with a standard deviation. Software R version 3.5.1 (R Development Core Team, 2010) was used to determine treatment effects and the standard error of these treatment effects. We converted all pairwise comparisons for a given study into contrast-based data for our analysis. The frequentist approach exploits the analogy between treatment networks and electrical networks to construct a network meta-analysis model accounting for the correlated treatment effects.^8,9 Due to expected heterogeneity between studies, we decided on reporting results from a random-effects model and this was an a priori decision.

The initial search yielded 394 articles. After full-text review, 10 RCTs with 1370 randomized patients were selected for analysis.^{10,,,,,,,,–19}Table 1 shows the characteristics of the trials and participants. Figure 1(a) summarizes the result that compared with control both beta-blockers and ARBs resulted in a similar and significantly reduced ARG mm/year at a mean difference of –0.39 (95% confidence interval (CI) –0.62, –0.16) and –0.42 (95% CI –0.65, –0.20) respectively, whereas the most significant reduction in the ARG was noted with the use of a combination of beta-blocker and ARB, leading to a mean difference of –1.16 mm/year (95% CI –1.53, –0.78). Figure 1(b) and (c) confirms these findings, showing that the combination of beta-blocker with ARB results in a significantly lower ARG compared with a beta-blocker, with a mean difference of –0.74 mm/year (95% CI –1.03, –0.44), and compared with an ARB, with a mean difference of –0.76 mm/year (95% CI –1.05, –0.47).

Figure 1.

Forest plot according to network meta-analysis of change in aortic root growth with different therapies.

MD: mean difference; CI: confidence interval; ARB: angiotensin receptor blocker; BB: beta-blocker.

Table 1.

Baseline characteristics of studies, participants and the interventions.

First author last name/ (study), year	Population (N)	Follow up (months)	Mean age (years)	Intervention/ comparator	Total daily dosage	Conclusion of the study	Study quality score
Tahernia, 1993	6	36	9.2	Intervention: beta-blocker Comparator: Control	Propranolol 30 mg	Beta-blocker better	++
Chiu, 2013	28	36	13.1	Intervention: beta-blocker Comparator: ARB–BB	Atenolol 50 mg or propranolol 40 mg Losartan 50–100 mg	ARB–BB combination better (p = 0.03)	++++
Groenink, 2013 (COMPARE)	145	36	37.6	Intervention: ARB Comparator: control	Losartan 50–100 mg	ARB better (p = 0.02)	++++
Sandor, 2014	17	12	17.3	Intervention: ARB Comparator: beta-blocker	Losartan 25 mg Atenolol 25–50 mg	No difference	++
Lacro, 2014 (Pediatric heart network)	535	36	11.3	Intervention: ARB Comparator: beta-blocker	Losartan 0.4–1.4 mg/kg Atenolol 4 mg/kg	No difference	++++
Bhatt, 2015	33	6	35	Intervention: ARB Comparator: beta-blocker	Losartan 100 mg Atenolol 50 mg	No difference	+++
Milleron, 2015 (Marfan Sartan)	299	36	29.9	Intervention: ARB Comparator: beta-blocker	Losartan 50–100 mg Atenolol 65 mg daily (mean)	No difference	++++
Forteza, 2016	113	36	25.2	Intervention: ARB Comparator: beta-blocker	Losartan 25–50 mg Atenolol 25–50 mg	No difference	++++
Muiño-Mosquera, 2017 (Ghent Marfan)	22	36	36.1	Intervention: ARB–BB Comparator: beta-blocker	Losartan 50–100 mg	No difference	++++
Mullen, 2018 (AIMS)	172	36	6–40a	Intervention: ARB Comparator: beta-blocker	N/A	ARB better (p < 0.01)	++b

a

Range.

b

Based on abstract only – full publication is pending.

BB: beta-blocker; ARB: angiotensin receptor blocker.

Aortic root dilatation is considered a prognostic marker as an increasing AGR heralds a poor outcome.²⁰ Our study is the first one to suggest that a combination of ARB and beta-blocker may lead to better outcomes compared with a single agent. The mechanism by which beta-blockers and ARBs may affect AGR is different and, therefore, it is theoretically plausible that a combination of both agents might work better.

Our study has several noteworthy limitations, and the results of our analysis should be considered hypothesis-generating. First, there is significant heterogeneity amongst trials included due to the design differences of the included studies. Also, the percentage of pediatric population differed widely between the included trials. Furthermore, the dosage of ARB used was trivial compared with the effective dose that showed promising results in a mouse model. Even though AGR is a surrogate marker of aortic dissection there is a lack of clinical outcome data from these trials. Due to a lack of large trials because of lack of participation and early terminations, we think that this is the best way to analyze the available data. Finally, even though our analysis shows benefits in prescribing the combination of ARB and beta-blocker together, the real-world application of this remains challenging. The side effects and pill burden associated with two medications are not assessed in this analysis, and they should be considered when deciding on appropriate treatment.

In conclusion, our analysis shows beta-blockers and ARBs as equally efficacious for prevention of aortic root dilatation. ARB should be added to the treatment regimen if tolerated by the patient as the combination of ARB–beta-blocker results in the most effective prevention of aortic root dilatation. More randomized studies are needed to evaluate the combined effects of ARB and beta-blocker on the progression of aortic root dilatation.

Acknowledgements

PROSPERO registration: CRD42019119753. Part of this research was presented as a moderated poster at American College of Cardiology Annual meeting 2019 at New Orleans, Louisiana, USA.

Author contribution

AM contributed to the conception or design of the work. SY, GP, SM, and WA contributed to the acquisition, analysis, or interpretation of data for the work; and drafted the manuscript. All authors critically revised the manuscript. All gave final approval and agree to be accountable for all aspects of work ensuring integrity and accuracy.”

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.