The atrial uremic cardiomyopathy regression in patients after kidney transplantation – the prospective echocardiographic study

Basic laboratory data prior to transplantation and over a 3-year follow-up are included in Table 1. Changes of creatinine level and eGFR over 3-years of observation in patients before and post-KTx are graphically shown respectively in Figs. 1 and 2.

All patients were treated with the immunosuppresion regimen, which included prednisone, mycophenolate mofetil and cyclosporine microemulsion or tacrolimus. All patients throughout the observation period received prednisone and mycophenolate mofetil (Table 1). Immediately after the KTx, half of the patients were treated with cyclosporine and the other half tacrolimus (Table 1). During the follow-up, the number of patients receiving cyclosporine in favour of tacrolimus decreased (Table 1).

Patients were also receiving other medicines than immunosuppressants such as antihypertensive drugs, taken both before and after KTx. After KTx, the use of antihypertensive drugs, especially diuretics, was reduced in many patients, however, the number of patients who received lipid-lowering treatment increased. Treatment prior to KTx and after is described in Table 1.

Detailed data on planimetric and volume parameters of LA in the studied group are presented in the second column of Table 2. Baseline characteristics of the hemodynamic indices of LA in the studied group is described in the second column of Table 3. Baseline characteristics of the hemodynamic indices of LV in the studied group is described in the second column of Table 4.

A detailed analysis of the changes in the LA planimetric parameters calculated using Student’s t-test has showed a progressive reduction in dimensions (Table 2). The reduction was found to be biphasic with a faster reduction in the first period of observation and then the rate of size reduction has been released. The volumetric indices followed the described changes confirming a significant reduction in the LA enlargement (Table 2).

There was observed a constant statistically significant reduction in values of LA planimetric parameters from 3 months after KTx until end of observation. The reduction in LA dimensions was fastest soon after KTx and limited in the latter stages for parameters LA_max, LA_min and LA_vaweP. However, among other parameters such as LA_circmax and LA_areamax the downward trend was almost stopped in the final stage of observation (Table 2).

Similarly as LA volumetric parameters, both the LAV_max and LAVI values constantly and statistically significantly decreased during the study period (Fig. 3). The LAV_max initially was rapidly decreasing in the first year after KTx and then the reduction was getting slower, but remained statistically significant. LAVI reduction showed a similar pattern with a faster decrease in the first year after KTx than subsequently (Fig. 4). Table 4 shows the changes concerning planimetric, volumetric, hemodynamic and mass parameters of LV observed during the 3-year follow-up.

There were very interesting changes of the LA hemodynamic indices. A gradual reduction in the LA dimensions was accompanied with an improvement in its function. In addition to the progressive reduction in the LA chamber size, the contractility also has improved, which was manifested in a gradual and statistically significant increase of LA_EF, LA_AE, LA_IE and LA_FS (Table 3), while the value of LA_PE was gradually decreasing (Fig. 5). However, in the case of LA_FS the most statistically significant improvement was made due to the changes between the 3rd and 12th months post-KTx, contrary to a still significant but a much smaller improvement recorded in the last observation period (Fig. 6).

Using the Spearman’s rank correlation a detailed analysis of LAVI with the tested echocardiographic parameters was carried out. Among the many echocardiographic indices obtained just after KTx significant correlation with LAVI and LA parameters as follows: LA_max, LA_min, LA_waveP, LA_shortmax, LA_longmax, LA_circmax, LA_areamax, LAV_max, LAV_min, LAV_waveP, LA_EF, LA_AE, LA_IE and LA_FS was observed (Table 5). Significant relationships between LAVI and LV indices just after KTx included: LVSV, LVM and LVMI (Table 5). The analysis of parameters obtained at the end of a 3-year follow-up showed a statistically significant relationship between LAVI and LA parameters as follows: LA_max, LA_min, LA_waveP, LA_shortmax, LA_longmax, LA_circmax, LA_areamax, LAV_max, LAV_min, LAV_waveP, LA_EF, LA_IE and LA_FS was noted (Table 6). In addition, the following significant relationships were found between LAVI and LV parameters: LVEDd, LVESd, PWDd, LVSV and FS (Table 6).

Multivariate analysis performed by multiple regression revealed that among the echocardiographic parameters analyzed immediately after KTx, LAV, LA_FS, LVM and LVMI were independently associated with LAVI (Table 7). The assessment made after the 3-years follow-up showed that LA_shortmax, LA_longmax, LA_circmax LAV_max, LA_FS, LVEDd, LVESd, PWDd and LVSV were independently associated with LAVI (Table 7).

This study demonstrated statistically significant reductions in the LA planimetric and volumetric parameters during the 36 months of observation. The reduction was found to be biphasic with a faster reduction in the first period of observation and then the rate of change slowed down. However, the constant reduction of dimensions brought very significant differences at the end of the observation. LAVI is a marker of LV diastolic dysfunction, hence its reduction is a proof of improvement in LV diastolic function during the study. The LA volume reduction cannot be considered to be secondary only to one pathophysiological mechanism. ESRD leads to structural and functional changes known as uraemic cardiomyopathy [23]. The pathogenesis of these disturbances is complex and so far the explanation is not completely understood. Among the potential factors determining the occurrence of these changes is fluid retention [24], arterial hypertension, accumulation of uraemic toxins, anaemia, hyperparathyroidism [25], renal malnutrition-inflammation complex syndrome [26], electrolyte disorders, interstitial fibrosis [25], disorders of the autonomic nervous system, as well as the presence of an arterio-venous fistula [27]. The changes in the circulatory system, including cardiac alternations resulting from uraemia can be divided into two groups. Firstly, the hemodynamic changes due to the uraemia appear and secondly serious metabolic changes accompanying this complex pathology undergo. These two basic pathogenic paths mutually coexist and a precise investigation of all the possible connections is impossible.

The observed biphasic changes in LAVI values show certain regularity. Probably an early reduction in LAVI value is more dependent on changing hemodynamic conditions as a result of reducing the volume overload accompanying caused by ESRD overhydration. Such observation derive from the study of Barberato et al. [10] in hemodialyzed patients. In the current study we observed that a sudden improvement in hemodynamic parameters and especially reducing the preload due to rapid and substantial dehydration caused a significant reduction in LA_max as well as volume parameters LAV and LAVI. Further late reduction of LAVI value recorded in present study should be rather considered as the resolve of uremia related toxemia and its negative impact on the remodeling of both LV and LA. This does not preclude the fact, that the removal of adversely affect negative hemodynamic factors stimulating structural changes in the heart cavities, plays a significant role in this slow recovery.

The LA changes towards restoration of its structure and function after the period of unfavourable remodelling is defined as reverse remodelling. A gradual reduction in LAVI indicates this phenomenon. To the best of our knowledge, the LAVI data post KTx reported in this study is unique. Previous publications did not examine the data in the same way as this paper. The present study shows a gradual change in LA dimensions proving the existence of reverse remodelling. To date, available reports only compare the data of patients after KTx with healthy persons at a certain time point rather than describing hemodynamic changes over time. The paper by Rysz et al. [28] describes the LA dimensions in patients 30 months post procedure compared with healthy individuals and showed a persistent significant increase of LA_max and LA_waveP. The study by Sahagún-Sánchez et al. [29] did not show any changes in the LA_max dimension after 3-month follow-up observation post-KTx. It is not known however, whether the period of observation was too short for change to occur. In both papers above the LA_max dimension in the study patients basically remained normal as it was 36.9 ± 4.5 mm and 35.9 ± 4.4 mm, respectively, whereas in the group of patients after KTx analysed in present study the starting dimension of LA_max was 41.64 ± 4.93. Thus the lack of improvement in the above studies compared to this study could be due to the differences in starting LA_max. A reduction from normal cavity size dimensions is more difficult to prove. On the basis of the above studies, KTx appears to contribute to favourable remodelling and restoration of the heart function. It is also known that even in dialysed patients there is as a short-term reduction of LA_max from 39.4 ± 8.1 mm to 37.8 ± 8.0 mm during haemodialysis. Also follow-up observations of patients after KTx who were continuing to have a preserved fistula compared to those, in whose fistula was removed, has revealed a significant difference in LA dimensions. The studies by Cridlig et al. [27] showed that LA_areamax in patients with an active fistula is higher (19.3 ± 4.8 cm²) than in patients without a fistula (16.4 ± 3.8 cm²). Moreover, this study showed for the first time that after a reduction or removal of ESRD driven triggers of reverse remodelling more normal heart dimensions is restored.

The undoubted advantage and novelty of the current research is the fact that these studies allow us to trace the mechanical renewal process of LA as a result of a dramatic change in hemodynamic and toxo-biochemical conditions. Previous data regarding the mechanical function of LA after KTx are scarce and limited to the paper by Rysz et al. [28].

The LA mechanical function can be described by three phases within the cardiac cycle [30]. Firstly, as a reservoir, the LA stores pulmonary venous return during LV contraction and isovolumetric relaxation (related to LA_EF, LA_IE). Secondly, as a conduit, the LA during the early phase of ventricular diastole transfers blood passively into the LV (related to LA_PE). Thirdly, the “contractile” function of the LA normally serves to augment the LV stroke volume between by approximately 15 and 30% (related to LA_AE) [30, 31].

Detailed analysis of the cardiac parameters shows a consistent improvement in hemodynamic atrial function such as LA_EF, LA_AE and LA_IE during the 3-months post-transplant and the improvement continues until the end of the observation period. In contrast, LA_FS was increasing relatively slowly but it resulted in a significantly higher value by the end of the observation, even though at earlier time points post KTx the changes did not reach statistical significance. Such behaviour of the LA functional parameters probably results from the fact, that indices such as LA_EF, LA_AE, LA_PE, LA_IE are more dependent on hemodynamic changes, which occur quickly and enable them to improve, while LA_FS as a parameter relating to myocardial contractility improves more slowly, which is probably caused by the chronic adverse effect of uremic toxins on the structure of the atrial tissue and the function of myocytes. It makes the time necessary for perceptible improvement in their contractility longer.

LA remodelling being a manifestation of atrial uremic cardiomyopathy, seems to be an example of complex cardiomyopathy. It consists of elements of both dilated and infiltrative cardiomopathy. There are no reports in the literature regarding this phenomenon in relation to LA. Analogies can be sought in relation to ventricular uremic cardiomyopathy as well as in reports of atrial cardiomyopathies studied in other clinical issues. And so LA_EF reduction occurs in patients with idiopathic congestive cardiomyopathy, where also a LA enlargement appears, what is known as primary atrial cardiomyopathy (atriomyopathy) [32]. Another mechanism of LA deterioration is associated with cardiomyopathy depending on pressure and volume overload as well as impaired LA function secondary ischemia, where the LA_EF decrease is originally a consequence of the increase in atrial afterload [33]. Histological evidence of fibrosis of the walls of the LA are in fact more pronounced in the case of dilated cardiomyopathy compared to small changes associated with myocardial infarction, which suggests that primary LA abnormalities in dilated cardiomyopathy should not be considered solely as a consequence of mechanical overload [34]. The observation that exercise capacity is directly dependent on the LA_EF value, while LA_EF value negatively correlates with the size of LA, emphasize the importance of LA as a pump for functional capacity of patients with dilated cardiomyopathy [35]. A significant decrease in LA systolic function and kinetic energy transferred to LV is also characteristic of patients with amyloidosis, as a result of infiltration of atrial myocardium by amyloid [36]. However, before the echocardiography visible the atrial wall infiltration occurs LA_EF remains within the normal range [36]. Both in dilated and infiltration cardiomyopathy there is also a deterioration of the LA reservoir function, increase in pulmonary artery pressure which affects the development of right heart failure [31]. Experimental models of atrial cardiomyopathy provide an insight into the mechanical consequences of LA systolic dysfunction, which can mimic the characteristics of dilated and infiltrative cardiomyopathy. Rapid atrial pacing (400/min) for 1 week in dogs caused global and regional deterioration of LA contractility while not contributing significantly LV function [37]. An increase in the E/A ratio and a reduction in the S/D ratio recorded from pulmonary flow were associated with a simultaneous decrease in LA_EF and an increase in LA’s conduit functions. However, maintaining fast pacing over a longer period of time (6 weeks) resulted in a further deterioration of LA compliance, deterioration of its diastolic function, impairment of its reservoir function and increase the proportion of the conduit function, so changes characteristic for clinical conditions associated with pressure and volume overload of LA [38]. Interestingly, LA failure caused by rapid stimulation has little or no effect for cardiac output and right ventricular function. It was provided that the LV function remained normal, because the increase in conduit function of LA was compensated by the reduction of LAEF and deterioration of the reservoir LA component [38, 39]. However, in the case of impaired LV diastolic function, the increased conduit function of LA is not able to compensate an impaired LA systolic function and reduced LA reservoir capacity [39]. The experimentally obtained data can therefore be particularly important in cases of dilated or infiltrative cardiomyopathy, where the LV systolic and diastolic dysfunctions are also often present [31]. The changes described on LA can also be used to depict LA changes in patients with ESRD and explain phenomena observed during the regression uremic atrial cardiomyopathy, which undoubtedly combines the elements of overload cardiomyopathy depending on unfavourable hemodynamic factors with elements of infiltrative cardiomyopathy associated with uremic toxins.

The current results are also important for future research on patients with ESRD who are scheduled for KTx. Two pathophysiological factors useful for guiding patients in the future are important. First, the use of monitoring the size of the LA (especially LAVI) for the assessment of LV diastolic function. LAVI reflects the long-term exposure of LA to elevated LV filling pressure. LAVI is a valuable parameter enables long-term hemodynamic monitoring of patients. By analogy, it can be stated that LAVI is in HF this same as glycated hemoglobin in monitoring patients with diabetes. This is particularly important and very useful in patients with volume overload that can accompany kidney disorders including monitoring the process of organ rejection. When qualifying for KTx, we evaluate mainly the LV systolic function (EF). Diastolic function is usually ignored. This is unsuitable because of the possibility of HF with preserved EF (HFpEF). HFpEF are highly prevalent in patients with ESRD. And this form of HF also have prognostic significance in patients on heamodialysis and being planned to KTx [40].

Secondly, the size of the LA is important in the context of the risk of atrial fibrillation (vicious circle: LA enlargement - atrial fibrillation and vice versa - reduction of the size of LA after KTx - lower risk of arrhythmia). This is fact of particular importance in the context of potential anticoagulant therapy, including the currently preferred therapy with new oral anticoagulants (NOAC). These drugs require more careful monitoring of renal function in order to avoid bleeding complications.

Prospective studies are now required to validate our results and define the LA structure and function impact on overall patient morbidity and mortality and – ideally – its reversibility with medical treatment and/or KTx.

We accept that this study has some limitations. The study was conducted on a relatively small number of patients using echocardiography. However, considering the nature of recruiting patients and data from the literature, the studied group is a relatively large population. This is due to the fact that the kidney transplantations are relatively rare in Poland. In addition, the analysis excluded patients who for various reasons did not undergo a full 3-year follow-up. A long 3-year observation, unprecedented in the literature in this type of research, is a major logistical challenge due to the fact, that patients enrolled in the study at the local transplant center, for obvious reasons, came from across the country. This fully explains and compensates for the relatively small size of the group.

The impact of the causes of ESRD were not analyzed because the group was too small to analyze the impact of the etiology on echocardiographic parameters (eg. there was only 1 patient with chronic tubulo-interstitial nephritis, and in 4 cases the etiology was not known). Statistical analysis with such a small number of patients in the subgroups does not make sense.

Pre-transplant echocardiographic information has not been given because they were not complete and available for all patients. Transplantations for obvious reasons were performed urgently. Recipient arriving for KTx often comes outside of our center, from all over the country. Performed echocardiography examination has covered only the basic parameters, and especially has not stated the parameters relating to the LA. Therefore, it could not be included in the study. As soon as possible after the HTx (after a satisfactory post-operative wound healing in order to avoid complications) echocardiography was performed according to a uniform protocol prepared for this project. Practically, it took place within a few days after the KTx. It can therefore be assumed that the parameters just before KTx did not differ significantly immediately after surgery. Uniform study protocol is also essential for the quality of follow-up data. In accordance with current orders, detailed assessment of the LA is not necessary prior to the KTx. The study also did not analyze the parameters of the LA in terms of prognostic. This was not the purpose of the work. The analysis included only those patients who survived the entire period of observation and have a good function of the transplanted organ.

Another limitation is the lack of analysis of the relationship between biochemical parameters and echocardiographic parameters. This study, however, was purely echocardiography, as highlighted in the title. Its advantage is undoubtedly the progressive character of the observation. The study also did not take into account the influence of applied pharmacotherapy, which may potentially affect the function of the LV and indirectly the function of the LA. This refers to antihypertensive agents, in particular ACE-inhibitors and immunosuppressive drugs. This was abandoned because ACE-inhibitors were not used, the effect of other antihypertensive drugs on the LA function is negligible. While the application of immunosuppressive therapy were similar in the whole group so its potential interference parameters in the LA was without significant influence on the entire test group.