Abstract
Objectives:
Arterial hypertension (AH) may be related to fluid retention, increased vascular resistance or hyperdynamic heart function. Impedance cardiography (ICG) is shown to be useful in the individualization of antihypertensive therapy but little is known about who most benefits from this therapeutic approach. The aim of this analysis was to estimate the effectiveness of ICG-guided antihypertensive therapy with respect to baseline blood pressure (BP) from the perspective of 12 weeks’ observation in randomized, prospective and controlled trials.
Methods:
This analysis involved 272 patients (average age: 44.1 ± 10.8 years) with AH. After baseline evaluation, including: office BP measurement (systolic, SBP; diastolic, DBP; mean, MBP) and ambulatory BP monitoring (mean 24-h SBP, mean 24-h DBP) the subjects were randomly assigned to groups of empiric [GE] and ICG-guided antihypertensive therapy [HD]. The results were evaluated separately in subgroups derived from median of MBP (110 mmHg): with slightly increased (‘SI_BP’) and more increased BP (‘MI_BP’). The comparative analysis included absolute change in BP (d_OSBP, d_ODBP, d_24-h SBP, d_24-h DBP) and the percentage of patients with reduction of BP ⩾ 10 mmHg (d10_OSBP, d10_ODBP, d10_24-h SBP, d10_24-h DBP).
Results:
ICG-guided therapy was shown to be superior to the empiric approach, especially in MI_BP. In this subgroup, the BP reduction in HD was higher than in GE: d_OSBP (23.3 ± 10.8 versus 18.5 ± 13.9 mmHg; p = 0.035), d_ODBP (16.0 ± 6.3 versus 11.6 ± 9.6 mmHg; p = 0.003), d_24-h SBP (17.7 ± 10.8 versus 13.1 ± 13.1 mmHg; p = 0.035). This benefit was also confirmed by a higher percentage of patients with significant BP reduction: d10_OSBP (87.7% versus 69.1%; p = 0.012), d10_ODBP (69.2% versus 47.3%; p = 0.012) and d10_24-h SBP (72.3% versus 52.7%; p = 0.012). The comparison in the SI_BP subgroup did not reveal such significant differences.
Conclusions:
The hemodynamically guided pharmacotherapy results in greater BP reduction. This effect is more pronounced in patients with higher baseline BP, while in those with slightly increased BP the empiric approach seems comparable to ICG.
Keywords
Introduction
Arterial hypertension (AH) is the main cardiovascular risk factor, which affects about a quarter of the world’s population. Although the number of hypotensive drugs is still increasing, only a small number of hypertensives achieve satisfactory blood pressure (BP) control [Kerney et al. 2005; Mancia et al. 2013]. Current guidelines [Mancia et al. 2013; James et al. 2014] clearly emphasize that hypotensive therapy should be individualized. However, the potential to assess hemodynamic profiles by standard clinical examination is significantly limited. Consequently, there is a need for diagnostic tools that will help in the identification of the complex mechanisms of increased BP, i.e. hyperdynamic heart function, fluid retention or increased vascular resistance.
Impedance cardiography (ICG), an easy noninvasive technique of hemodynamic assessment, is shown to be effective in the individualization of hypotensive treatment. The pharmacotherapy utilized for hemodynamic alterations such as increased resting heart rate (HR), cardiac index (CI), high thoracic fluid content (TFC) and elevated systemic vascular resistance index (SVRI) is shown to be more beneficial than the empiric treatment in several randomized and prospective trials [Ventura et al. 2005; Taler et al. 2002; Smith et al. 2006; Krzesiński et al. 2013]. Thanks to those observations, ICG is perceived as prophylactic support in the selection of hypotensive drugs. However, there is some doubt as to its efficacy in patients with slightly increased BP, compared with those requiring two and more medications because of high BP [Taler, 2014]. Moreover, in view of limited reimbursement for medical care, it is important to determine who benefits more from hemodynamically guided individual therapy.
Therefore, the aim of this analysis was to estimate the effectiveness of ICG-guided antihypertensive therapy with respect to baseline BP from 12 weeks’ observation in randomized, prospective and controlled trials.
Methods
Study population
The study group consisted of patients with at least a three-month history of AH defined according to the European Society of Cardiology (ESC) guidelines [Mancia et al. 2013]. Exclusion criteria were: (1) confirmed secondary AH; (2) AH treated with three or more medicines before recruitment; (3) significant heart failure; (4) cardiomyopathy; (5) significant heart rhythm disorders; (6) significant valvular disease; (7) kidney failure (GFR below 60 ml/min/1.73 m2); (8) chronic obstructive pulmonary disease; (9) diabetes; (10) polyneuropathy; (11) peripheral vascular disease; and (12) age < 18 years.
The group selected for this analysis comprised subsets of patients from two prospective, randomized and controlled studies, performed in the Department of Cardiology and Internal Diseases of the Military Institute of Medicine. The first included 128 subjects (registered at http://www.nauka-polska.pl, ID: 227062) and was conducted between 2008 and 2009 and the second included 144 subjects [ClinicalTrials.gov identifier: NCT01996085] and was conducted between 2012 and 2014.
The studies were conducted according to the Good Clinical Practice guidelines and the Declaration of Helsinki, with the approval of the local ethics committee. Each patient provided written informed consent to participate in the study.
Office blood pressure measurement
Office blood pressure measurement (OBPM) was measured in the morning hours (07:30–08:30) using automatically operated equipment (Omron M4 Plus, Japan), a technique compliant with the European Society of Cardiology guidelines [Mancia et al. 2013]. Office systolic blood pressure (OSBP) and diastolic blood pressure (ODBP) were measured in a quiet room, in the presence of a trained physician or nurse following a minimum of 5 minutes’ rest in a sitting position. The BP category of AH was defined according to the ESC guidelines [Kearney et al. 2005]: grade 1, an OSBP 140–159 mmHg with or without an ODBP 90–99 mmHg; grade 2, an OSBP 160–179 mmHg with or without an ODBP 100–109 mmHg; grade 3, an OSBP greater than 180 mmHg with or without an ODBP greater than 110 mmHg.
Ambulatory blood pressure monitoring
Ambulatory blood pressure monitoring (ABPM) was commenced during morning hours (Spacelabs 90207, Spacelabs, Medical Inc., Redmond, WA, USA). The time period 06:00–22:00 was considered the daily activity period (daytime) with automatic BP measurement at 10-minute intervals. During night rest (night-time: 22:00–06:00) the measurements were performed every 30 minutes. The patients were advised to adjust their circadian activity to these periods of time. The minimum correctness of BP measurement was defined as 70% for both daytime and night-time. The BP thresholds used to define AH were set according to the ESC guidelines [Mancia et al. 2013]: a mean 24-hour SBP ⩾ 130 mmHg with or without a DBP ⩾ 80 mmHg, a mean daytime SBP ⩾ 135 mmHg with or without a DBP ⩾ 85 mmHg, or a mean night-time SBP ⩾ 120 mmHg with or without a DBP ⩾ 70 mmHg. The values of mean 24-hour SBP and DBP were included in this analysis.
Impedance cardiography
All ICG measurements were performed using the Niccomo™ device (Medis, Ilmenau, Germany) after 10 minutes of rest in a supine position. BP measurement was performed automatically every 2 minutes with an arm cuff. The other hemodynamic parameters were measured by beat-to-beat method. The values of TFC, CI, SVRI and HR from the fifth minute of the examination were taken into account in the treatment algorithm. The cut-off values of TFC, SVRI, CI and HR (from the fifth minute of the ICG examination) defined the hemodynamic profile: (1) hyperconstrictive in the case of SVRI > 2500 dyn*s*m2/m5 (where ‘*’ means multiple); (2) hyperdynamic if CI > 4.2 l/min/m2 with or without HR > 80/min; (3) hypervolemic profile if TFC > 34 1/kOhm for men and >24 1/kOhm for women; (4) balanced profile if hemodynamic parameters below established threshold values. The subjects requiring combined therapy with regard to significantly increased BP (average office BP > 160/100 mmHg with or without average 24-h BP > 140/90 mmHg) were also identified.
Study design
Both studies were randomized (1:1), prospective and simultaneously controlled by conventional treatment. Initial clinical evaluation was performed face to face or by a phone call. Patients who had been taking medicines before the study were advised to stop and in the case of increased BP, to take captopril sublingually (a minimum of 7 days of pharmacological ‘wash-out’).
On the second visit, all patients underwent the complete clinical examination, performed with consideration of cardiovascular risk factors and symptoms indicating secondary cause of AH [Mancia et al. 2013]: interview and physical examination, OBPM, ABPM, ICG, electrocardiogram, echocardiography and laboratory tests (ions, creatinine, fasting glucose and lipids).
Following examination, the patients were randomized to two groups: (1) empiric (GE) or (2) hemodynamic (HD) group. Figure 1 shows the protocol flowchart (summary of the two patient subsets) for observation of mean time 94.2 ± 10.3 days. There were 31 patients excluded from the final analysis because they met the exclusion criteria: 18 resigned from control visit; 6 decided to discontinue recommended therapy within the first 2 weeks; 3 patients were excluded from pharmacotherapy because of borderline BP (nonpharmacological treatment recommended); and 4 met exclusion criteria after randomization. The final analysis comprised 241 subjects.
The study flowchart (summary of two cohorts of patients).
Independent researchers made the treatment choice in both groups. Each researcher was blinded to the assigned treatment on the visit after 12 weeks and evaluated treatment effects (i.e. including OBPM and ABPM). Differences between the groups with regards BP reduction and obtained BP control after 12-week treatment were considered as final points (as per protocol analysis).
Treatment
Nonpharmacological treatment was administered according to the current ESC guidelines [Mancia et al. 2013]. Pharmacotherapy included drugs where effectiveness in hypertension treatment had been previously confirmed [Mancia et al. 2013]: lisinopril (ACEI, angiotensin converting enzyme inhibitor), telmisartan (ARB, angiotensin receptor blocker), hydrochlorothiazide/indapamide (diuretic), metoprolol/nebivolol (BB, beta blocker) and amlodipine (CB, calcium blocker).
The treatment algorithm in the HD group, arbitrarily predetermined and described in detail in our previous study [Krzesiński et al. 2013], was based on our own data collected in a cohort of hypertensive patients, and the analysis of the previous reports [Taler et al. 2002; Smith et al. 2006; Sramek et al. 1996; Flack, 2006]. The different cut-off values of TFC for men and women were defined referring to the sex-dependent computational algorithm used in a Niccomo device. Cut-offs for CI and SVRI were adopted from the previous studies of Taler and colleagues [Taler et al. 2002], as well as Smith and colleagues [Smith et al. 2006]. Increased resting heart rate (RHR) greater than 80 beats/min was considered as an indication to use a beta-blocker because it had been reported to be an unfavorable prognosis [Palatini, 2007; Palatini et al. 2006].
In brief, the first-line drug choice was based on the hemodynamic profile defined as aforementioned. In the case of a hyperdynamic profile, BB was recommended; a hypervolemic profile suggested diuretic treatment; and a hyperconstrictive profile meant treatment with vasodilator(s). Combined therapy was applied in the case of complex hemodynamic disturbances and in patients requiring combined polytherapy (BP criteria mentioned above).
The effectiveness of ICG-guided therapy was evaluated in subgroups derived from baseline median of office mean BP (110 mmHg) of: slightly increased (‘SI_BP’: n = 121), and more increased BP (‘MI_BP’: n = 120).
Statistical analysis
Statistical analysis was performed with the use of Statistica 7.0 software (StatSoft Inc., USA). Normality of data distribution was checked by visual inspection, as well as using the Shapiro–Wilk test. All results were expressed as average values ± SD for continuous variables and number of subjects (percentages) for categorical ones.
Merging the subjects from two trials allowed the researchers to maintain compliance with the terms of sample size calculation (described previously [Krzesiński et al. 2013]) for both derived subgroups. The comparison between GE and HD subjects was performed independently in the SI_BP and MI_BP subgroups. The treatment effects for OBPM and 24-hour ABPM in those subgroups were compared with the use of analysis of variance (ANOVA)/Mann–Whitney U-test for continuous variables (change in BP), and Chi squared test/Fisher’s exact test for categorical variables (percentage of achieved reduction in BP of minimum 10 mmHg, and percentage of BP control after 12-week treatment). A p value of less than 0.05 was considered statistically significant.
Results
Baseline clinical data
The final analysis involved a group of 241 patients (167 men) of average age 44.1 ± 10.8 years (range from 19 to 68 years), whose baseline characteristics have been presented in Table 1. No significant differences were observed between the GE and HD subjects in the whole study group, and the SI_BP and MI_BP groups.
Basic characteristics: data presented as mean ± SD and numbers (percentages).
AH, arterial hypertension; BMI, body mass index; DBP, diastolic blood pressure; SBP, systolic blood pressure; GE, empiric group; HD, hemodynamic group; SI_BP, slightly increased blood pressure; MI_BP, more increased blood pressure.
Treatment effects
Hemodynamically guided therapy resulted in higher BP reduction both in OBPM and ABPM (Figure 2). The analysis in subgroups revealed that its beneficial effect in the whole group (Figure 2A) was mostly the derivative of significant BP decrease in MI_BP subjects (Figure 2C). The comparison in SI_BP subgroup did not reveal significant differences between the empiric and the hemodynamic approach (Figure 2B).
Comparison of the effect between empiric (GE) and hemodynamically guided treatment (HD) within BP change (p value presented on the bars) in: the whole group (A), SI_BP (B) and MI_BP (C).
The benefit of ICG in MI_BP was also confirmed by a higher percentage of patients who presented with BP reduction of a minimum of 10 mmHg (Figure 3C). Hypotensive therapy adjusted to hemodynamic profile was also more effective in reaching appointed BP control (Table 2), again as a result of the relevant clinical advantage in the MI_BP subgroup.
Comparison of the effect between empiric (GE) and hemodynamically guided treatment (HD) within achieved reduction in BP of minimum 10 mm Hg (p value presented on the bars) in: the whole group (A), SI_BP (B) and MI_BP (C).
Treatment effect within BP: data presented as numbers (percentages).
BP, blood pressure; DBP, diastolic blood pressure; OBPM, office blood pressure measurement; ODBP, office diastolic blood pressure; OSBP, office systolic blood pressure; SBP, systolic blood pressure; GE, empiric group; HD, hemodynamic group; SI_BP, slightly increased blood pressure; MI_BP, more increased blood pressure.
Pharmacotherapy
The most commonly administered drugs in both groups were the renin–angiotensin–aldosterone (RAA) system’s blockers and diuretics (Table 3). In the whole group, diuretics and BB were more frequently used in the HD subgroup, but at the same time GE patients received more RAA blockers. SI_BP/HD subjects were those that contributed most for the diuretic advantage in the HD subgroup, while in the MI_BP/HD group the differences in pharmacotherapy were not statistically significant. However, the trends in more frequent recommendation for BB and CB were observed.
Pharmacotherapy: data presented as numbers (percentages).
RAA, renin–angiotensin–aldosterone; ACEI, angiotensin converting enzyme inhibitor; ARB, angiotensin receptor blocker; BB, beta-blocker; CB, calcium blocker; GE, empiric group; HD, hemodynamic group; SI_BP, slightly increased blood pressure; MI_BP, more increased blood pressure.
Discussion
Hypertension control remains a challenge for clinical cardiologists [Mancia et al. 2013; Ferrario et al. 2010]. According to the current guidelines, the management of AH ought to be focused on detailed diagnostics, but the tools to support the selection of individually adjusted antihypertensive polytherapy are still limited [Mancia et al. 2013; James et al. 2014].
In the present analysis we aimed at verifying whether the beneficial effect of the ICG-based algorithm of antihypertensive therapy proposed in our previous study [Krzesiński et al. 2013] is dependent on baseline BP. We observed that the proposed strategy of treatment provided additional value in subjects with a higher BP. In those with only a slightly increased BP it was not superior to the empiric approach. These observations suggest clinically important conclusions: first, the benefit of ICG increases with severity of AH; and, second, the use of ICG in patients with mild AH would not significantly influence the BP reduction related to hypotensive therapy.
The analysis in the whole study group confirmed that the use of hemodynamic measurements increases the effectiveness of antihypertensive management. The patients treated according to ICG achieved greater BP reduction in both OBPM and ABPM and higher rate of final BP control in OBPM. Even statistically nonsignificant differences are clinically important in terms of prognosis because a decrease in BP of only 4/3 mmHg is associated with a lower risk of stroke (by 23%), coronary disease (15%), and overall mortality (14%) [Williams, 2005]. The results of comparisons in subgroups (SI_BP and MI_BP) addressed the main aim of the analysis and showed that subjects with higher BP are the true beneficiaries of ICG.
The effect of hemodynamically guided therapy in MI_BP subgroup was comparable with that observed in the CONTROL study [Smith et al. 2006]. The authors evaluated 164 patients with essential AH (average BP 155/93 mmHg) and obtained better office BP control (<140/90 mmHg) in HD group (77% versus 57%; p < 0.001). The marginally better BP control than in our MI_BP subgroup (69% versus 43.6%; p = 0.003) might be explained by more the complex methodology of intervention in the CONTROL study, which included several additional visits during 3 months’ follow up and more changes of pharmacotherapy (mean of 1.1 new drugs introduced per treatment cycle). In another prospective, randomized and controlled trial, Taler and colleagues used ICG-based therapy in a series of 104 resistant hypertensives (average BP 171/89 mmHg) and also achieved higher frequency of goal office BP in HD group (56% versus 33% in empiric group, p < 0.05) [Taler et al. 2002]. Fadl Elmula and colleagues confronted the BP-lowering effect of pharmacotherapy adjusted to the individual hemodynamic profile with renal denervation in a small sample of patients with AH (n = 19) [Fadl Elmula et al. 2014]. After careful selection of true treatment-resistant subjects, nine of them were randomized to invasive treatment and ten to the subgroup treated according to the hemodynamic alterations identified by ICG. The patients in that subgroup started with comparable office BP (160/88 mmHg), as in the case of our MI_BP subjects (153/100 mmHg). They achieved slightly lower than that of our MI_BP’s BP reduction after 3 months (20/7 mmHg versus 23/16 mmHg), but after the next 3 months the effect was improved (28/11 mmHg). Most importantly, the ICG-guided therapy was shown to be more effective than renal denervation.
The analysis of the pharmacotherapy applied in our study revealed the tendency to more intensive treatment in the HD subgroup. However, significant differences were observed only with diuretics and BB in the whole group, and diuretics in the SI_BP subgroup (the medications less frequently recommended than RAA blockers). In the MI_BP subgroup it was only the tendency to more frequent use of CB and BB, resulting mostly from hemodynamic indications to polytherapy with RAA blocker. It suggests that in the case of SI_BP we can achieve comparable treatment effects using different medicines, even in monotherapy, and the use of ICG would not significantly influence BP reduction. Conversely, in the MI_BP, the benefit of ICG appears as a possibility to adjust combined therapy according to the individual hemodynamic profile.
Limitations
Slightly more intensive pharmacotherapy in the HD subgroup could influence final BP but we should emphasize that the drug selection was based on a predefined, researcher-independent, two-step algorithm and the evaluation of the effect was blinded. On the other hand, there is no ground to hypothesize that the treatment in GE, recommended by an experienced researcher, was in its assumption less aggressive. In our opinion the significant difference in the hypotensive effect between the HD and GE groups overcomes this potential bias.
We are aware that the other study limitations are the sample size, which influences the statistical power of the comparison in subgroups, and post hoc division into SI_BP and MI_BP, based on the arbitrary BP cut-off value (median office MBP). It should be also noted that we recruited hypertensives without other serious chronic diseases. Thus, our results should not be extrapolated to the general population.
Conclusion
ICG is a useful diagnostic method, which offers the additional option of individualized and effective drug selection in AH. The hemodynamically guided pharmacotherapy effects are observed with a greater BP reduction in patients with higher baseline BP. However, in those with mild AH the empiric approach seems to be comparable to ICG. It suggests that the potential utility of ICG increases with the greater complexity of AH and pharmacotherapy. It should be considered in the choice of potential candidates for ICG-guided therapy.
Footnotes
Acknowledgements
We would like to thank the medical staff of the Department of Cardiology and Internal Diseases of Military Institute of Medicine, especially Professor Andrzej Skrobowski for the assistance in patient care and organizational supervision, and Dr Beata Uziębło-Życzkowska for the assistance in patient care, performing echocardiography and data collection. In addition, we are grateful to Dr Robert Wierzbowski, Dr Jarosław Kowal, Dr Małgorzata Kurpaska, Dr Katarzyna Hałas, Dr Magdalena Potapowicz-Krysztofiak, Dr Agnieszka Jaguś-Jamioła, Dr Łukasz Michalczyk, Dr Agnieszka Wójcik, Dr Anna Kazimierczak, Dr Agnieszka Jurek, Dr Kalina Wolszczak and Dr Agata Galas for the assistance in patient care and data collection. We also thank Małgorzata Banak for the assistance in ICG measurements, as well as Lidia Wojda and Lidia Latosek for their nursing care and data collection.
Funding
This work was supported by Ministry of Science and Higher Education/Military Institute of Medicine, Warsaw, Poland (grant number 148/WIM).
Conflict of interest statement
The authors declare that there is no conflict of interest.