This study is first evaluating RCA flow velocity pattern and flow reserve in patients with HF resulting from advanced non-ischemic cardiomyopathy, employing invasive spectral Doppler technique. We showed that in these patients RCA phasic flow pattern has still a mild diastolic predominance, contrarily from verified experimentally [
7‐
9], and that was not significantly influenced by occurrence of pulmonary hypertension or RV dysfunction (Figs.
1,
2,
3,
4). Aside, we have no observed any significant correlation with echocardiographic and hemodynamic parameters with RCA flow pattern or RCA-CFR (Table
1). Moreover, the RCA flow reserve did not differ from LAD, even when pulmonary hypertension or RV dysfunction were present (Fig
5, Table
3).
Coronary phasic flow pattern
Our findings showed a mild diastolic predominance in RCA flow pattern (Figs.
1 and
2), differently from some animal experiments also in face of hemodynamic abnormalities, in which it was observed a systolic flow preponderance [
7‐
9]. Actually, in dogs RCA is small and limited to RV [
13], so its branches do not receive influence from vigorous LV contraction over intramural resistance vessels during systole, which is considered a main responsible in attenuating systolic flow [
11].
In our study, the LAD diastolic flow predominance was marked, compatible with expected LV mechanics influence (Figs.
1 and
2). In fact, Krams et al. have shown experimentally that the LV contractility or elastance were most responsible to restricting systolic coronary flow, even compared to intracavitary pressure [
12,
34]. Also in normal human being it was observed a systolic flow attenuation in LAD [
14]. Akasaka et al. studying patients with hypertrophic cardiomiopathy [
35], as well as Yoshikawa et al. studying patients with aortic stenosis [
36], have shown a still more distinct impediment in systolic component, or even a reverse systolic flow.
RCA dominance is present in majority of humans [
29], as it was observed in our patients, so once RCA branches reach and penetrate LV myocardium, we can assume that LV muscle contraction should also influence in any degree the RCA phasic flow pattern, explaining part of our findings. Possibly, in a small RCA marginal branch, just restrict to RV and not representative of entire RCA ramifications, and particularly in a more distal coronary analysis, we could find a diverse coronary flow pattern, as observed by Okura et al. [
37] and others investigators [
38,
39] studying another cardiopathies and circumstances.
Right cardiac side has several structural and functional differences from left cardiac side [
3‐
6]. For example, RV free wall is generally thinner than LV wall, in normal and in patients with dilated cardiomyopathy [
4,
6]. Therefore, we could infer that also in human being, the RV intraventricular pressure, or even the RV wall stress, could have still more influence in the mechanics of coronary flow pattern [
3,
6]. Nevertheless, also in patients with pulmonary hypertension or with RV structural and functional repercussions, we have not verified any effect in RCA flow pattern. Although such influence have ever been demonstrated in animals – as in Lowensohn's study [
9], particularly analyzing an extremely increased RV intracavitary pressure – these acute and non-physiological experimental conditions could not represent the chronic compensated human HF state, even under pulmonary hypertension condition. The significant RV dysfunction – which represent the final common pathway of diverse structural and mechanical RV abnormalities [
3,
4] – have not influenced significantly the RCA flow pattern, in comparison to preserved RV functional subgroup (Fig.
4). Other than, the complex RV structure may have influence from LV disarrangement [
3,
4,
6], aside from the dysfunctional interventricular septum. It is inferably that in this chronic and stable advanced cardiomyopathy HF setting, these disturbances in right cardiac side were not robust enough to exacerbate RCA systolic flow attenuation. For those same reasons we attributed the absence of significant correlation among any isolated echocardiographic and hemodynamic parameters and RCA phasic flow pattern, as described on Table
2. So, the RCA balanced contact with both RV and LV, and their respective simultaneous structural and hemodynamic contributions, probably is one of the main reasons for this lacking of any robust correlation.
Coronary flow reserve
Independently of inherent mechanical and hemodynamic differences between right and left cardiac sides in advanced heart failure due to dilated non-ischemic cardiomyopathy [
2‐
4,
6,
40], the CFR was equivalent in RCA and left coronary circulation (Fig.
5). In settings without such deterioration, i.e. in normal human beings, Ofilli et al. have reported a balanced RCA versus LAD flow reserve, employing alike technique [
14]. The several vascular intrinsic and extrinsic mechanisms enrolled in preserving the CFR (19) possibly were in account to promote our findings. Inclusive, the former mechanism has surely a diffuse nature in the cardiac vessels, as already it was verified experimentally [
41].
We have employed adenosine, to provoke hyperemia, that is fundamentally an endothelium-independent stimulant and acts at arteriolar level [
31]. Thus, instead of analyze intrinsic factors, we have assessed mostly the resistance vessels, that are particularly under mechanical extrinsical influences [
19,
42]. For sure, LV structural disturbances – that include since wall and cavity compromising until perivascular disarrangement – had any kind of effect on CFR [
19]. Additionally to any RV compromising level, certainly the RCA extension to LV, prominent in RCA dominance circumstances, had an important role to maintaining that CFR equivalence. Probably, these are also the explanations for the lacking of significant correlation of RCA-CFR with any echocardiographic or hemodynamic variable isolatedly, once their effects were consequently minimized (Table
4). These findings are contrastable to the assessment of LAD-CFR and the correlation with LV parameters exclusively, as verified by Inoue et al. [
16].
The presence of pulmonary arterial hypertension has also a prognostic role in advanced HF patients [
27]. When we have compared the RCA to LAD flow reserve in the pulmonary hypertension patients group, we did not find a significant difference (Table
3). Fixler et al, have found a decrease in coronary flow, and a subsequent RV failure, in setting of markedly high RV intraventricular pressure, in dogs [
10]. As a matter of fact, at initials levels of RV hypertension, the RCA flow even increases, corresponding to a vasomotor adaptation and reserve, including in acute conditions [
10,
43]. Only when a more extreme RV systolic pressure is reached, this vasomotor reserve exhausts, then occurring that hemodynamic consequences [
10]. Nevertheless, Manohar et al. had observed experimentally in ponies that even in presence of an elevated right intraventricular pressure, and a consequent marked driving pressure reduction, the RCA flow reserve was preserved [
13]. This RCA-CFR capability to overcome hemodynamic adversities was observed experimentally by Murakami et al. as well, employing also adenosine in dogs [
44]. Certainly, this evidenced capability range of CFR adaptation – added to RCA interaction to both RV and LV, as observed in humans and also in ponies [
13] – and the, not experimental, chronic and relatively clinical stable conditions found in our patients, have contributed to our balanced findings. Other than, the not significant heart rate variation during the stress, in our patients, also could have a role in RCA-CFR preservation, as remarked in Manohar's study [
13]. In RV dysfunctional circumstances, we also did not find a difference between RCA and LAD flow reserve (Table
3). The RV functional assessment in this HF setting was also appealing due to particular prognostic implications [
1,
2,
4]. However, the presence of significant RV dysfunction have not influenced the RCA-CFR enough to make a distinction with the left arterial CFR. As a matter of fact, there are many different reasons influencing RV systolic function [
3‐
5], particularly in advanced NIC scenario, making any correlation with RV function difficult. Some of such reasons could be the LV-RV muscle contiguity, ventricular interdependence, direct injuries to RV myocardial fibers, or even the high pulmonary artery pressure
per se [
3,
4]. Apart from that, the ejection fraction, as an ejection phase index, is influenced by load conditions, hence limiting the effective myocardium fiber contraction analysis [
3‐
5]. It is possible also, that a presence of any kind of vasomotion corresponding interaction in LAD as RCA be stressed [
43]. Therefore, aside from the different mechanisms in preserving CFR, it is reasonable to accept the weak relationship found in this study enrolling RV dysfunction over RCA versus LAD CFR, in conditions of advanced chronic non-ischemic heart failure syndrome.
Although it has not been object of the present study, the absolute increment verified on CFR merits some considerations. Firstly, we had no control group to compare with and thus to endorse these findings. Secondly, even though we had analyzed advanced non-ischemic cardiomyopathies, the different etiologies could have a role in our findings. In fact, few studies have evaluated CFR in Chagas' disease. Torres et al. have found a marked reduction of CFR with acetylcholine, not paralleled with the adenosine infusion [
18]. We have also observed a preserved adenosine CFR in another study, but assessing Chagas' heart disease with no LV dysfunction [
45]. Some older studies assessing CRF in dilated cardiomyopathies have found a diverse CFR impairment comparing endothelium with non-endothelium stimulation [
17,
19]. Treasure et al. analyzing CFR in dilated cardiomyopathy have found a significant reduction under acetylcholine stimulus, but it had no difference comparing with controls under adenosine provocation [
17]. By other hand, Bitar et al have reported a variable response on coronary arteries after acetylcholine infusion in non-ischemic dilated cardiomyopathy HF setting [
46]. Vanderheyden et al. have verified a blunted adenosine CFR in idiopathic dilated cardiomyopathy, but they speculated the higher resting coronary flow as one of the reasons for these findings [
47]. Nevertheless, most of recent studies accords with the non endothelium-dependent (i.e., adenosine and dipyridamole) CFR reduction in dilated non-ischemic cardiomyopathy setting, principally studying LAD territory – justified mainly due LV structural and mechanics deterioration aside the microvascular disarrangements, as described by Rigo et al. and others [
16,
19,
47,
48]. Thirdly, it is worth of note that we have used a more incisive adenosine protocol to looking for the highest hyperemic peak. Fourthly, it is known that the CFR in this HF setting could present individually some diversity, maybe being lower in most advanced cases, or even more related with HF functional class, as reported by Santagata et al. [
49]. And finally, as commented by Nitenberg and Antony, it has a lot of variables influencing the CFR assessment, considering the different techniques, protocols and studied population, therefore demanding meticulousness in comparative analysis from different studies [
50].
Clinical implications
This study demonstrated how is the RCA flow behavior in conditions of HF in advanced non-ischemic dilated cardiomyopathy. Although the RCA phasic flow pattern had been different from that evidenced in experimental studies [
7,
9], we met similar findings than described in normal human being [
14,
15]. Certainly, these findings are in accordance with the several mechanisms of coronary flow adaptation in human being [
19]. Other than, it is possible to suppose that any process involved in the diastolic function improvement also could ameliorate the predominant diastolic coronary flow, in not only the left but also in the right cardiac side, in this HF setting [
51,
52]. Hence, minimizing the compromised systolic component, and then creating a favorable vicious circle, inclusive with a benign repercussion on ventricular systolic function. These aspects could have even therapeutic implications. Also, it is noteworthy that it is still technically difficult to get some information non-invasively from RCA [
53], mainly detailed spectral Doppler curves to diastolic/systolic proportion analyses, making primary, at the moment, to employ invasive studies to improve these human RCA physiopathology understanding.
Respecting CFR, by these results we can infer that is possible to extrapolate the findings from LAD to RCA, or vice versa, in this setting of patients. Generally it is easier to explore the CFR in left coronary circulation (LAD) due to its topography [
16‐
19], even non-invasively [
48,
53,
54], avoiding time-consuming and the technical limitations concerning RCA circulation assessment. This information has more importance when one wishes to evaluate the global non-regional microcirculation, in a given non-ischemic cardiomyopathy population, employing invasive or non-invasive Doppler techniques [
18,
19,
48,
53,
54]. Rigo et al have recently demonstrated the prognostic importance of CFR, employing transthoracic dipyridamole echocardiography at left coronary artery territory in patients with dilated NIC [
48]. Such study have permitted inclusive to recommend particular treatment strategies based on CFR behavior [
48]. Our study could endorse the diffuse nature of CFR findings at chronic dilated NIC setting, relatively independent of right cardiac repercussions. Apart this, to assert that efforts to improve RCA-CFR could have a role in benefiting RV failure in advanced NIC or even in survival, it will require more investigations.