The challenge of early glomerular filtration rate decline in response to antihypertensive treatment and chronic kidney disease outcomes

Abstract

Hypertension and chronic kidney disease (CKD) are closely linked pathological processes. Combating high blood pressure (BP) is an essential part of preventing CKD progression and reducing cardiovascular (CV) risk. Data from recent randomized controlled trials on patients at high CV risk showed the beneficial effects of intensive action to meet BP targets on mortality related to CV disease. The impact of meeting such targets on renal function is still unclear, however, particularly for patients with CKD. This issue has been the object of several post hoc analyses because lowering BP definitely has a nephroprotective role, but the early decline in glomerular filtration rate (GFR) associated with antihypertensive therapies and strict BP targets is still a concern in nephrology clinical practice. The present review discusses the results of studies on this topic, focusing specifically on the clinical significance of early GFR decline in response to treatment with angiotensin-converting enzyme inhibitor/angiotensin receptor blocker, or to different BP targets, in terms of renal and CV outcomes, and how this tips the balance towards continuing or discontinuing antihypertensive therapy.

INTRODUCTION

Chronic kidney disease (CKD) and systemic hypertension are closely connected. Fluid overload, salt retention, activation of the renin–angiotensin–aldosterone system and sympathetic overstimulation increase as kidney function gradually declines, leading to the onset of hypertension [1]. High blood pressure (BP), on the other hand, is the main cause of progressive kidney function loss [2]; it disrupts the autoregulation mechanisms of glomerular filtration, inducing the remodelling of afferent arterioles and nephro-angiosclerosis [3]. Interrupting this vicious circle is the cornerstone of efforts to slow CKD progression and reduce cardiovascular (CV) risk and related death rates [4]. Which BP values should be targeted in CKD patients is still a matter of debate, however, and the guidelines differ not a little regarding this indication [5].

The American Heart Association and American College of Cardiology guidelines released on November 2017 suggest a <130/80 mmHg BP target for CKD patients, irrespective of their degree of proteinuria [6]. The 2018 European Society of Cardiology/European Society of Hypertension Guidelines [7] recommend lowering office-measured systolic BP to values in the range of 130–140 mmHg for patients with diabetic or non-diabetic CKD. These guidelines also stress the need for tailored therapeutic strategies, taking tolerability, electrolyte disorders and renal function into account. Several meta-analyses recently showed a reduction in all-cause mortality for CKD patients under strict BP intervention, suggesting a significant benefit of pursuing this approach for such patients [8–10]. The 2012 Kidney Disease: Improving Global Outcomes (KDIGO) guidelines on BP management in patients with CKD were revisited at the Controversies Conference in September 2017, the recently published conclusions of which do not go against the latest recommended BP targets [11]. The authors concluded that, while intensive BP control seems to be safe in patients with CKD, recent data from clinical trials should be carefully reviewed before any new guidelines can be established for such patients. An important concern is the early loss of renal function associated with the use of angiotensin-converting enzyme inhibitors (ACEIs)/angiotensin receptor blockers (ARBs) and strict BP control. This is because of a strong association between a 30% and 40% decline in estimated glomerular filtration rate (eGFR) during the first 12 months of treatment and the subsequent reaching of an established endpoint for CKD progression [i.e. a doubling of serum creatinine (SCr)] was reported in a meta-analysis of 37 randomized controlled trials (RCTs) [12]. So, the question is whether the early GFR decline that occurs within 6 months of initiating strict antihypertensive treatment has the same dismal significance as a (possibly more gradual) decline occurring over a 12-month follow-up. This question is addressed in the present narrative review.

BLOOD PRESSURE CONTROL AND LONG-TERM RENAL OUTCOME

While it is clear that CKD patients benefit from strict BP control in terms of all-cause and CV-related mortality risk [8–10], whether the same can be said as regards nephroprotection is controversial. In a meta-analysis of nine RCTs, Tsai et al. [9] analysed the relationship between the adoption of intense (<130/80 mmHg) or standard (>140/90 mmHg) BP targets and major renal outcomes, defined as a doubling of SCr or 50% reduction in GFR, end-stage kidney disease (ESKD), composite renal outcome and all-cause mortality. More than 8100 non-diabetic CKD patients of various ethnicities (African-American, Caucasian and Asian) were followed-up over a relatively short period (median 3.3 years). Compared with standard BP control, adopting more intensive BP targets brought no advantage in terms of renal protection. Only non-black individuals and those with higher levels of proteinuria showed a trend towards a lower risk of kidney disease progression with intensive BP control [9]. Heterogeneous case populations and short follow-ups may explain why no nephroprotection was demonstrated, however. In fact, a recent meta-analysis with a longer follow-up (14.9 years) of the African American Study of Kidney Disease and Hypertension (AASK) and Modification of Diet in Renal Disease (MDRD) studies highlighted a significant ESKD risk reduction for some patients on intensive BP intervention, including those ≥40 years old, those with proteinuria ≥0.44 g/g creatinine and those with a body mass index ≥30 kg/m² [13]. Taken together, these data suggest that intensive BP control may have a different nephroprotective effect in different subsets of CKD patients.

EARLY GFR LOSS INDUCED BY BP MEDICATION

As mentioned previously, the early decline in renal function associated with the administration of antihypertensive drugs is a matter of concern, particularly when this effect is more severe. A decrease in GFR soon after the initiation of an intensive BP control regimen is a common finding, especially with ACEI/ARB [14–17]. The haemodynamic changes associated with antihypertensive therapy may prompt a drop in renal blood flow (RBF) and intraglomerular pressure, producing an early reduction in GFR. This loss of function may be a price worth paying in exchange for a lower CV risk [18] and/or a more long-term kidney protection, given that a lower intraglomerular pressure attenuates the vascular damage responsible for nephrosclerosis [19]. As regards renal protection, uncertainty about the safety of such a drop in GFR stems from the fact that renal hypoperfusion may lead to acute kidney injury and to chronic damage in the form of renal fibrosis and ESKD [8].

In the trials considered in this article, the generally mild reduction observed in GFR (of the order of no more than 5–7% of its baseline value) seems to be safe. This was elegantly demonstrated by Apperloo et al. [15] already in the 1990s, in a small group of 40 non-diabetic, hypertensive patients with CKD (eGFR 90–30 mL/min/1.73 m²) given double-blinded treatment with enalapril or atenolol and followed-up for 4 years. The study aimed to highlight the role of early eGFR reduction in CKD progression and the reversibility of this initial eGFR loss. An initial reduction in GFR occurred in both the randomized groups, with no significant differences between them. Irrespective of the treatment administered, patients with a distinct initial reduction in eGFR (averaging ∼7% within 3 months) showed a gentler slope in their kidney function loss than patients with no drop in eGFR. After the treatment was withdrawn at the end of the follow-up, renal function was restored (almost to pre-treatment eGFR levels) in the group with a larger early reduction in GFR, whatever the treatment administered, while this did not happen in the group with a smaller early GFR decline.

These observations suggest that the early, minor eGFR losses occurring after the establishment of antihypertensive treatment have a functional origin and are potentially nephroprotective [15]. A very similar mechanism has recently been proposed to explain the analogous effect on GFR induced by sodium-glucose co-transporter-2 (SGLT2) inhibitors [20]. The meaning of major early reductions in GFR (merGFR) may be different and is still unclear.

THE CONCEPTS OF MAJOR EARLY REDUCTIONS IN GFR AND INCIDENT CKD AFTER THE SPRINT TRIAL

Several post hoc analyses of RCTs on large study populations and two retrospective studies on general practice cohorts have recently investigated this matter (Tables 1 and 2).

Two ways to reduce the ‘merGFR’ occurring briefly after starting BP treatment to an irrelevant clinical problem have been proposed, one focusing on the statistical phenomenon of regression to the mean [17, 24], the other on seeking acute kidney injury biomarkers in patients manifesting merGFR [29, 30]. Relying on regression to the mean, the assumption is that patients initially disclosing an extreme merGFR will tend towards a value nearer the average GFR when they are retested. We disagree with such an interpretation, because treatments likely to have modified renal haemodynamics had already been adopted, as also stressed by Weir [31]. Two studies investigated urinary biomarkers in Systolic Blood Pressure Intervention Trial (SPRINT) patients with ‘incident CKD’ and concluded that this condition was due not to renal injury but to a haemodynamic phenomenon [29, 30]. We find these studies inconclusive since rather heterogeneous conditions were included in the ‘incident CKD’ group, making its definition difficult. In fact, the authors’ analysis did not distinguish between patients with early- or late-onset ‘incident CKD’, and those whose ‘incident CKD’ did or did not recover.

In our opinion, that the definition of ‘incident CKD’ is troubling is the most important issue to consider in all those trials that used it as a surrogate renal outcome (Table 1). In SPRINT patients with a normal baseline renal function, ‘incident CKD’ was defined as a >30% decline in GFR to a value <60 mL/min/1.73 m² (confirmed by a subsequent test ≥90 days later) whenever it occurred during the entire follow-up (average 3.1 years) [31]. But this is not an unequivocal demonstration of CKD because it does not rule out the possibility of the drop in GFR fall being a purely haemodynamic phenomenon. In the sub-analysis of the SPRINT trial on individuals with a normal baseline renal function, there were 180 patients—in both the intensive and the standard treatment arms (3.7% and 1%, respectively)—who developed ‘incident CKD’, 98 during the first 6 months of follow-up and 82 afterwards. During the follow-up, they partially or totally recovered from this ‘CKD’ in 66 of 180 cases (37%), suggesting a functional rather than structural effect. Although none of these patients developed ESKD (with the caveat that the follow-up was relatively short at a median 3.1 years), 17 had a further decline in GFR to beyond 50%, corresponding approximately to a doubling of SCr [29]. Unfortunately, we are not told whether these 17 patients’ GFR declined gradually from the beginning or the >50% decrease in GFR was associated with their ‘incident CKD’ during the first 6 months of treatment [32].

As outcomes, ‘ESKD’ or ‘a doubling of SCr’ would be much more robust as evidence of CKD progression, but the trials listed in Table 1 were not long enough or sufficiently powered to investigate differences in such outcomes. Data obtained from post hoc analyses of the available trials provide information on the cumulative incidence of CKD during the follow-up, without distinguishing between cases occurring early as opposed to later on. This distinction seems important because the pathophysiological and clinical meaning of incident CKD could differ when it develops soon after starting intensive antihypertensive treatment or at a later point during a patient’s follow-up. An ‘incident CKD’ occurring later in the follow-up might actually be just the sign of progression of a chronic nephropathy. A further limitation of the analysis conducted on these trials lies in that the rates of partial or total recovery from ‘incident CKD’ are not taken into account (with the notable exception of the SPRINT trial).

A post hoc analysis of the Action to Control Cardiovascular Risk in Diabetes Blood Pressure (ACCORD BP) trial was performed specifically to shed light on the renal effects of an intensive as opposed to a normal BP control regimen, borrowing the SPRINT definition of ‘incident CKD’ [33]. Unlike the SPRINT trial, however, which did not include diabetic patients, the ACCORD BP only enrolled patients with Type 2 diabetes mellitus (DM2). The incidence of CKD was found higher than in the SPRINT trial (at 3-year follow-up, it was 10% in the intensive intervention arm and 4.1% with the standard regimen) [33]. The analysis did not go into much detail, however; in particular, it failed to consider the rates of partial or total recovery from ‘incident CKD’, as in the SPRINT trial. Interestingly, the incidence of CKD was significantly higher among DM2 patients with albuminuria, suggesting that incident cases of CKD in such patients could be genuinely non-haemodynamic, chronic conditions.

Peralta et al. [22] recently ran a post hoc analysis on RCTs on the secondary prevention of small subcortical stroke to assess the effectiveness of two antiplatelet treatments and two BP targets (<130 mmHg versus 130–149 mmHg) in preventing stroke among patients with a previous lacunar stroke. Only 16% of the population enrolled had a GFR <60 mL/min/1.73 m². SCr was measured yearly. Neither primary nor secondary renal outcomes were preset because the trial was not designed or powered to identify the onset of CKD or ESKD. The authors found that both antihypertensive regimens induced a modest decline in eGFR during the first year after randomization, which was steeper in the intensive BP target group (−4% versus −1%) [22]. Renal function remained stable thereafter in both arms. The authors also defined patients whose GFR decreased by >30% as ‘rapid decliners’; this happened in 9% of patients in the first year (a very relative ‘early’ rapid decline!). Although the frequency of incident CKD was ascertained (different from the definition used in the SPRINT trial), it was impossible to see whether a predictive relationship existed between rapid GFR decline and ESKD because no data were available on the latter outcome.

Evidence of merGFR (>30% in the first 3–6 months after starting antihypertensive therapy) being safe, and often associated with a recovery of GFR, comes from a post hoc analysis of the Cardiovascular Outcomes in Renal Atherosclerotic Lesions trial in which patients with atherosclerotic renal artery stenosis (>60%) received optimal medical therapy or underwent stenting [23]. The analysis only addressed the results obtained in the ‘no stenting with optimal medical therapy’ group. A rapid decline (as early as 6 months after starting treatment) was observed in 18% of patients, but their GFR improved by a mean 15% in the ensuing 1-year follow-up. A few patients’ GFR recovered completely within 2 years after its initial decline. Only three patients developed ESKD and only one of them was from the ‘rapid decliners’ group. While the results of this study are generally reassuring as regards the long-term overall consequences of merGFR in patients with atherosclerotic renal artery stenosis, the possibility of a partial recovery of any initial GFR decline being due to some compensatory mechanism involving the contralateral kidney cannot be ruled out.

In short, we feel that the studies analysed thus far (Table 1) do not clearly demonstrate that an early >30% decline in GFR induced by treatment for hypertension in a non-trivial percentage of patients raises the risk of ESKD. This is because of criticalities in the surrogate endpoint for CKD.

MAJOR EARLY REDUCTIONS IN GFR INDUCED BY BP CONTROL AND ESKD

Six studies concerning eight trials/cohorts (Table 2) used strong clinical outcomes to address whether the merGFR induced by antihypertensive treatments or ACEI/ARB raised the risk of ESKD.

Holtkamp et al. confirmed the findings of Apperloo et al. regarding GFR slopes in the intervention arm (with losartan treatment) of the Reduction of Endpoints in DM2 with the Angiotensin II Antagonist Losartan cohort of CKD diabetic patients, which was much larger and had a longer follow-up [15, 17]. That said, patients with an early GFR decline in the third tertile (mean −8.3 mL/min/1.73 m², corresponding to a 20% decline) had significantly more renal outcomes (ESKD, creatinine doubling) than those with a smaller or no decline. In addition to losartan, patients in the intervention group could be given calcium channel blockers, β-blockers, centrally acting agents and diuretics, as necessary, however, and the interaction between different antihypertensive agents and the drop in GFR was not analysed.

In the post hoc analysis of 9340 patients involved in the Ongoing Telmisartan Alone and in combination with Ramipril Global Endpoint Trial and the Telmisartan Randomized Assessment study in ACE-intolerant participants with cardiovascular disease, with a median follow-up of 4.7 years, a >15% drop in GFR 8 weeks after starting ACEI/ARB was seen in 7% of patients and associated with a doubling of SCr or ESKD [hazard ratio (HR) = 4.27] [24].

In a post hoc analysis on the ADVANCE trial cohort of over 11 000 patients, an HR of 1.34 for new or worsening nephropathy was associated with an early >20% drop in GFR after starting antihypertensive treatment [25]. The outcome was defined as the onset of macro-albuminuria, a doubling of SCr to ≥200 µmol/L, or ESKD, after a mean follow-up of 4.4 years. In the active arm, treatment consisted of perindopril (2 mg) and indapamide (0.625 mg). All other treatments, except for ACEI and thiazide, could be continued at the discretion of the doctor in charge. No single class of antihypertensive agents was predictive of a rise in SCr levels.

The AASK and MDRD trials—with an extended follow-up of 11 and 6 years, respectively—were re-examined by Ku et al. to shed more light on the clinical meaning of the early decline in GFR after starting antihypertensive treatments [26]. Different classes of antihypertensive agents were used in both trials to reach the established BP targets. In the MDRD trial, the recommended antihypertensive regimen was an ACEI with or without a diuretic; a calcium channel blocker and other drugs were added as needed. In the AASK trial, patients were randomized to amlodipine, or ramipril, or metoprolol.

The majority of the patients enrolled were in stages CKD3b–CKD4. In both the strict and the standard BP control arms, an early drop in eGFR >20% in the first 3–4 months was associated with a higher risk of ESKD [26]. In the AASK cohort, there was no significant interaction between the randomized antihypertensive agents administered and the categories of percentage decline in eGFR [26].

Two real-life retrospective studies on huge general practice cohorts have investigated the issue. On over 122 000 patients retrieved from the UK’s Clinical Practice Research Datalink (which is linked to hospital records), it was shown that early SCr increases of >10% induced by ACEI/ARB were associated with a graded higher risk of CV endpoints, ESKD and mortality (the analysis was adjusted for comorbidities, including CKD stage, at the baseline) [27]. Patients were concomitantly taking any of the following antihypertensive drugs: β-blockers, calcium channel blockers, thiazides, loop diuretics and potassium-sparing diuretics. Loop diuretics and potassium-sparing diuretics were used four times more frequently by patients experiencing a >30% rise in SCr levels. A shortcoming of this study lies in the long interval (up to 14 months) between GFR measurements before and after initiating the treatment, since any differences identified could relate to long-term CKD progression rather than an acute decrease in GFR.

A recent retrospective study on almost 32 000 citizens of Stockholm accessing healthcare who started taking ACEI/ARB in the years 2007–11 found early rises in creatinine levels >10% increasingly associated with mortality, CV endpoints and ESKD [28]. This study definitely investigated acute variations in SCr levels because the interval between measurements was just 3 months.

MAJOR EARLY REDUCTIONS IN GFR INDUCED BY BP CONTROL, AND THE RELATED MORTALITY AND CV OUTCOMES

Table 3 presents that, in all the studies in which it was analysed, a merGFR (of >10, or 15 or 20%) raises the risk of mortality and CV endpoints. What is striking in the article by Fu et al. [28] is that, while only 10% of patients disclosing a GFR variation >30% actually developed ESKD, as many as 48% died or had a CV outcome during follow-up. Similar data emerged from the ADVANCE trial (4% versus 20%, respectively) [28] and in a British cohort (2% versus 45%, respectively) [27]. Clearly, there is a competing risk between CV outcomes and CV-related mortality on the one hand, and ESKD on the other, and the former prevails.

Interestingly, recent analysis found no evidence to justify any reconsideration of the benefit of intervention (ACEI/ARB or intensive BP control) on CV outcomes and mortality based on the extent of variation in GFR [18, 24, 25]. The issue with these analyses, however, is whether the lack of any effect in attenuating the beneficial effects of antihypertensive therapies was driven by the much larger subgroups experiencing little or no drop in GFR. In fact, only a tiny percentage of patients had merGFR. These trial-based studies were also post hoc analyses, and different criteria were used to define merGFR (in terms of the intervals between measurements, interval after starting the intervention and any distinction between persistent and transient GFR variations).

ARE MAJOR EARLY REDUCTIONS IN GFR INDUCED BY ANTIHYPERTENSIVE TREATMENTS CAUSALLY RELATED TO ADVERSE CV/RENAL OUTCOMES

We have seen that certain reductions in GFR induced by BP-lowering treatments may predict the risk of progression to ESKD, CV outcomes and mortality.

The gradual increase in the risk in clinical endpoints observed in some studies [25, 27, 28] supports the conviction that the magnitude of merGFR is a clinical problem rather than an expression of the statistical phenomenon of regression to the mean.

The higher risk was seen especially with ACEI/ARB and intensive antihypertensive treatments, but also in the control arm of the cited trials, that is, with standard BP control measures (probably because control patients involved in trials receive more attention, comply better with their treatments and have a target range of BP values to meet). The same applies whatever the stage of patients’ CKD and their background renal disease.

As a pathophysiological explanation for the increased risk, it has been claimed that the drop in GFR reflects existing renal and systemic vascular disease, and glomerular hyperfiltration—in other words, fragile renal haemodynamics [25]. We largely agree with these interpretations but would point to the elephant in the room, which is systemic BP. The fact that the phenomenon occurs when BP-lowering drugs are used leads us to suggest that patients have their own ‘critical’ BP capable of jeopardizing renal autoregulation and thus causing merGFR. For instance, the mean BP values reached in the intensive treatment arm in the SPRINT and other clinical trials are close to 80 mmHg (at the lower end of autoregulation in the normal kidney) [34]. In patients with longstanding hypertension, severe atherosclerosis or diabetes, the arterioles (and especially the afferent glomerular arterioles) undergo a process of hyalinosis. This interferes with the arterioles’ ability to adjust the diameter of their lumen to adapt their resistance to blood flow and cope with the systemic BP, keeping intraglomerular pressure and filtration constant. In other words, the lower end of the kidney’s autoregulation capacity could be >80 mmHg, and this is probably the case in a proportion of patients in the cohorts considered. This phenomenon is magnified by the use of ACEI/ARB because of their vasodilating effect on the efferent arterioles. The consequence of a further reduction in BP in such a circumstance is a drop in GFR and a lower renal blood perfusion, prompting renal ischaemia and triggering an unfavorable loop that is harmful to the kidney. Taking this view, the acute drop in GFR does not cause ESKD directly, but it does give rise to precarious renal haemodynamics that is ultimately harmful. An interesting experiment in a pig model of acute ischaemic nephropathy showed that medullary tissue hypoxia only occurs when RBF decreases by >20% (corresponding to a roughly similar drop in GFR) [35]. Hypoxia is known to induce interstitial inflammation and fibrosis, which, in turn, causes or aggravates renal damage [36]. This might explain how a ‘critical’ decline in GFR induced by antihypertensive treatment is associated with the progression of CKD. Smaller reductions in blood perfusion (identifiable from lower drops in GFR, generally <5%)—such as those seen in trials on patients with hypertension and diabetes, particularly when ACEI/ARB are investigated—do not induce renal ischaemia or interstitial damage. Instead, they facilitate a positive remodelling of the renal arterioles and glomerulus [34].

The fragile renal haemodynamics signalled by merGFR is likely to reflect other critical conditions in other vascular beds. We can therefore interpret merGFR as a marker of the risk of CV disease. This hypothesis is supported by the finding in the Swedish cohort [28] that the risk of certain outcomes—particularly overall mortality and CV endpoints—associated with a >30% increase in SCr was higher over a relatively short period (the first year of observation).

In our opinion, the J-shaped curve of the association for hypertension with mortality, CV-related and renal outcomes is the epidemiological counterpart of the same phenomenon [37].

SHOULD ANTIHYPERTENSIVE TREATMENTS THAT CAUSE MAJOR EARLY REDUCTIONS IN GFR BE DOWNGRADED, WITHDRAWN OR MODIFIED?

There is no denying that intensive BP control and/or treatment with ACEI/ARB helps to prevent CV and renal events in the majority of patients for whom they are indicated, but they may be contraindicated in some cases. A large early drop in GFR identifies the latter patients, who might be at high risk of subsequent adverse outcomes.

Until the findings of specifically designed studies clarify the picture, we think it is wise to mitigate antihypertensive treatments for patients showing a more marked initial drop in GFR. This recommendation is driven mostly by the CV risk to which such patients are exposed, which is far greater and earlier than the renal risk.

Since Clase et al. [24] found certain variability in GFR in the first 2 months after starting ACEI/ARB blockade, we suggest modifying or discontinuing such treatments only after multiple measurements obtained in the first 1–2 months have confirmed a significant drop in GFR.

We also think what is considered a critical decline in GFR should be adjusted to a patient’s basal eGFR. Both Holtkamp et al. [17] and Clase et al. [24] showed that the early drop in GFR is inversely proportional to the initial GFR. A drop of 10–20 mL/min/1.73 m² in CKD3 patients should therefore be given careful consideration, while the same attention should be paid to a drop >20 mL/min/1.73 m² in patients with a GFR >60 mL/min/1.73 m².

It might be safe to avoid or down-titrate drugs that interfere with RBF autoregulation (Table 4) [38]. As Tables 2 and 3 clearly show, early reductions in GFR capable of predicting the risk of ESKD and CV outcomes were investigated in trials and patient cohorts starting treatment with ACEI/ARB and in trials on intensive BP control regimens. To our knowledge, no such studies have considered antihypertensive drugs other than ACEI/ARB. This might be worth doing because the interaction of ACEI/ARB with other agents that interfere with glomerular haemodynamics could have a part to play. Renal autoregulation is governed by the myogenic response and tubuloglomerular feedback, and it is impaired by a number of conditions and drugs [38] used in most of the studies reviewed in this article (Table 2). In particular, we would recommend investigating calcium channel blockers and loop diuretics because they have a definite effect on renal autoregulation. Notably, the use of loop diuretics has reportedly been much more prevalent in patients revealing a marked drop in GFR [27].

CONCLUSIONS

MerGFR may occur in the early period of ACEI/ARB and antihypertensive treatments, particularly under intensive BP control regimens, both in individuals with a near-normal kidney function and in CKD patients. This condition seems to predict CV episodes, ESKD and mortality, though some authors would disagree. A final answer on this topic requires further studies, for instance a meta-analysis of cited studies to overcome the lack of statistical power of single trials.

In future trials investigating treatments that alter renal haemodynamics, sub-studies specifically addressing this topic should be planned. Important aspects to address in future trials at the aim of clarifying the topic are determination of the variability of patients’ GFR in a run-in period; determination of the rate of CKD progression prior to the trial; wash-out of the treatment and assessment of patients’ GFR at the end of follow-up; a clear and stated in advance definition of merGFR (interval between assays, interval between starting intervention and subsequent GFR measurements and persistent versus transient variations in GFR) and sharing of the merGFR definition by future trials to allow performing meta-analysis of results.

Until the results of such studies become available, we consider it reasonable to mitigate BP control/ACEI/ARB in patients disclosing merGFR.

CONFLICT OF INTEREST STATEMENT

The authors have no conflict of interest to declare. This article has not been published previously in whole or in part. This article was possible through a fund from the University of Verona.

REFERENCES