Effects of AMPD1 gene C34T polymorphism on cardiac index, blood pressure and prognosis in patients with cardiovascular diseases: a meta-analysis

Cardiovascular disease (CVD), a series of diseases involving the heart or blood vessels of people, has become one of the predominant causes of morbidity and mortality worldwide threatening human health according to W.H.O statistics [1‐3]. At present, CVD remains the first reason of great number of deaths and disabilities in China, and nearly 3 million individuals died each year from CVD, which approximately accounts for 42% of all deaths in 2010 [4]. In general, CVD, including coronary artery diseases (CAD), heart failure (HF), myocardial infarction and so on, are multifactorial disorders frequently sharing or coexisting risk elements like hypertension, dyslipidemia, overweight, and physical inactivity, which could partially illustrate the etiology of CVD [5, 6]. Apart from the above, mounting evidence has indicated that genetic polymorphisms in some special genes also have important impact on the pathogenesis of CVD, which provided solid foundations to elucidate the essence of clinical phenotype diversity and individual clinical treatment [7, 8].

The AMPD1 gene encodes the adenosine monophosphate deaminase (AMPD), a crucial enzyme in purine nucleotide and energy metabolism especially in skeletal muscle, as well as in cardiac muscle, with the function of catalyzing adenosine monophosphate to inosine monophosphate [9, 10]. To our knowledge, The AMPD1 gene is prevalently localized in chromosome 1 in the region p13 ~ 21 with 20 kb spanning in length, and consisting of 16 exons and 15 introns [11]. Under most conditions, there is a common polymorphism of the nonsense mutation with a C-to-T transition at position 34 of AMPD1, contributing to the appearance of a premature stop codon, which leads to some related metabolic muscle diseases due to the AMPD activity deficiency [12]. While several recent studies also suggested that AMPD1 may play essential roles in the initiation and development of CVD, for instance, it has been demonstrated that AMPD1 allele mutation play a protective role during congestive heart failure (CHF) development, which contributed to prolong possibility of survival without cardiac transplantation [13], meanwhile, it can alleviate the aortic stiffness and inflammation for patients with CAD [10]. Therefore, it was reasonably speculated that AMPD1 as a possible candidate gene is related to CVD. But the effect of AMPD1 C34T polymorphism in CVD remains controversial, like Collins RP and his team, who have yielded conflicting results stating that the mutant T allele of AMPD1was deleterious in both HF and post-MI patients who had myocardial infarction histories [14]. Besides, de Groote P et al. did not find any impacts of AMPD1 genetic mutation on the prognosis and survival in CHF patients [15]. Therefore, there appears a need for a meta-analysis to investigate the relationship between the AMPD1 gene C34T variant and CVD risk regarding the aspects of cardiac indexes, blood pressure and prognosis, and thereby deriving a more plausible estimation for CVD.

Our study showed that LVEF (%) in CVD patients with T allele of AMPD1 gene C34T polymorphism was significantly increased in comparison to those patients with CC genotype. On the contrary, the CVD patients with CT + TT genotype had a greater reduction of LVEDD (mm) than those with CC genotype. In addition, our subgroup analysis according to the patients with or without HF verified that the LVEF was markedly higher in HF patients who carried with CT + TT genotype than CC genotype. As is known to all, abnormalities of left ventricle (LV) function or structure has been a common event among most CVD patients, demonstrating a negative influence on cardiovascular prognosis [25, 26]. While LVEF is well-recognized as an important determinant of a person’s cardiac function to be clinically applied for the evaluation of the severity of heart systolic function, and in particular, it is typically low when the patients with the systolic form of congestive heart failure [27]. As for LVEDD, it has widely accepted as a sensitive indication for measurement of the left ventricular remodeling for patients with CVD [28]. More importantly, the myocardial remodeling, which is marked by changes in LVEF and LV volumes, could be functioned as a practical prognosis indicator to guide clinicians’ diagnose [29]. Probably, there is an explanation for this finding that this polymorphic variant of AMPD1 C34T could weaken the AMPD activity, and promote the elevated circulating levels of adenosine, a crucial protective agent, which was in part the consequence of the attenuation of myocardial fibrosis and ventricular remodeling, and thus contributing to less severe CVD [13, 30]. Consistently, both Safranow K et al. [21] and Gastmann A et al. [24] pointed out the tendency that LVEDD values in patients with CAD or CHF who carried the mutant T allele of AMPD1 was smaller, further indicating that the AMPD1 C34T polymorphism might affect different types of ventricular remodeling among CVD. Interestingly, no relevance of AMPD1 gene C34T polymorphism to LVEF (%) had been found in our 5 included researches [13‐15, 21, 24], which might be partially limited by the small sample size. To our knowledge, high blood pressure (BP), or together with other risk elements could maximize the cardiovascular risk to a certain extent [5]. In the present study, we also revealed that being carrier of the T allele in AMPD1 C34T polymorphism is remarkably correlated with lower levels of SBP, indicating that T allele carriers of this SNP may benefit for CVD individuals, as it have a decreased risk of developing the CVD pathologic state through a lower value of SBP. As shown by Safranow K, T allele carriers tended to have lower SBP than wild type in patients with CAD and HF, especially in ischemic HF patients, what might account for the lower fasting serum glucose or the higher creatinine clearance due to the T allele in patients with HF [21]. Moreover, fasting serum glucose has been conformed to be positively correlated with SBP in a previous research [31]. Therefore, the relatively increased LVEF and reduced LVEDD (or LV volumes) in CVD patients indirectly indicated that the T mutant allele of AMPD1 variant performed a protective effect on cardiac function recovery.

In addition, the present meta-analysis further explored the possibility of a relative survival and cardiovascular incidence benefit for the AMPD1 genetic mutation in HF patients, but failed to find any significant differences between different genotypes (CT + TT and CC) of this polymorphism and total survival rate and cardiac survival rate. It may therefore be possible that adenosine concentration in plasma (or muscle) of different genotypes showed no differences although T allele could induce the synthesis of myocardial adenosine [32, 33]. Moreover, the adenosine concentration in patients with T allele was only markedly increased after exercise as suggested previously, hence, the protective role of T allele might exert protective functions through increasing the local secretion of adenosine due to the short half-life of adenosine [34, 35]. Our result was broadly consistent with Kolek MJ group, who reported the AMPD1 genotype had no statistically influences on the survival of HF [23]. Similarly, no influence of the functional AMPD1 C34T polymorphism has been demonstrated by De Groote P et al. on the major clinical parameters and cardiac survival in consecutive white patients with stable CHF linked to left ventricular systolic dysfunction [15].

There are several limitations of the current meta-analysis need to be pointed out. First, as we know, heart failure can be caused by atherosclerotic and non-atherosclerotic factors, but we couldn’t conduct a subgroup analysis due to the fact that the included studies didn’t analyze the effects of AMPD1 gene C34T polymorphism on cardiac index, blood pressure and prognosis in patients with HF based on the etiology.. Second, heterogeneity in our study may influence the reliability of our results, although subgroup analysis was conducted to detect the heterogeneity source, as well as sensitivity analysis was introduced to appraise the stability of the results. Third, the data extracted from each record were based on unadjusted estimates, which may lead to misleading results. Fourth, we did not obtain all desired information from all materials because of the small number of studies. Fifth, the relatively small sample size of our study might have negative impact on the power of our meta-analysis. Lastly, the limited English and Chinese studies might also lead to language bias despite no evidence of publication bias from our statistical tests.