Follow‐Up of Cardiovascular Risk Markers in Hypertensive Patients Treated With Irbesartan: Results of the i‐SEARCH Plus Registry

Abstract

J Clin Hypertens (Greenwich).

Microalbuminuria (MAU), high‐sensitivity C‐reactive protein (hsCRP), and N‐terminal pro‐brain natriuretic peptide (NT‐proBNP) are risk markers used to predict the prognosis of hypertensive patients; however, they have not been prospectively evaluated in primary care. An investigation was conducted using i‐SEARCH Plus, a registry documenting 1649 patients with hypertension who received irbesartan at office‐based cardiologists over 12 months. Mean age at baseline was 61.4±11.3 years, 43.2% were women, and blood pressure was 159.8±20.1/93.4±11.9 mm Hg. Median albumin/creatinine ratio (ACR) at baseline was 9.90 (interquartile range [IQR], 5.76‐‐25.52) mg/g, hsCRP 2.46 (IQR, 1.16‐‐5.14) mg/L, and NT‐proBNP 89.28 (IQR, 38.63‐‐203.40) pg/mL. In patients with MAU (ACR ≥20 mg/g), the age‐adjusted risk of a combined end point of newly diagnosed coronary artery disease (CAD), myocardial infarction, stroke/transitory ischemic attack, and death at 12‐month follow‐up was increased (odds ratio [OR], 2.67; 95% confidence interval [CI], 1.49–4.76), as was the incidence of CAD (OR, 3.27; 95%CI, 1.39–7.68) and death (OR, 4.63; 95%CI, 1.44–14.94). No correlations with end points were found for hsCRP or NT‐proBNP after adjusting for age and the presence of MAU. MAU is an independent predictor of cardiovascular events in hypertensive patients. These findings confirm previous reports on the prognostic value of MAU and establish its incremental value over hsCRP and NT‐proBNP. J Clin Hypertens (Greenwich). 2010;12:909–916. © 2010 Wiley Periodicals, Inc.

Risk markers for future cardiovascular (CV) events gain importance in cases in which the absolute risk is low and the incremental efficacy of new therapeutic options is not dramatic. Hypertension in particular is a disease that gradually develops even at younger ages when CV events are infrequent. The determination of significant differences between drugs and drug classes in this patient population would require large studies lasting several years or even decades. It is important, however, to obtain these data because of the necessity of lifelong treatment.

Markers such as microalbuminuria (MAU), highly sensitive C‐reactive protein (hsCRP), and N‐terminal pro‐brain natriuretic peptide (NT‐proBNP) that correlate to the future development of CV disease have been identified from a number of epidemiologic studies. MAU in particular has even developed into a reliable indicator for end organ damage. This led to its recommendation for identifying high‐risk patients in the treatment of hypertension by the European Society of Hypertension (ESH) and the European Society of Cardiology (ESC). ¹ The presence of microalbuminuria in particular calls for the use of renin‐angiotensin system (RAS) blockers such as angiotensin‐converting enzyme (ACE) inhibitors and angiotensin receptor blockers (ARBs), which have been shown to have a lowering effect on these markers. ² , ³

The time course, the interdependence, and the clinical utility of these risk markers has so far, however, not been investigated in patients in clinical practice. It is unclear whether irbesartan, which has been shown to lessen the degree of microalbuminuria in type 2 diabetic patients in clinical trials, has impact on the predictive performance of these markers.

Therefore, we set up a prospective disease registry in which risk markers were determined at baseline in patients with arterial hypertension (≥140/90 mm Hg). Patients were prescribed irbesartan and followed for 12 months. The aim was to determine the risk of total mortality and CV events (newly diagnosed coronary artery disease [CAD], myocardial infarction [MI], unstable angina pectoris [UAP], and stroke/transient ischemic attack [TIA]) in relation to baseline values of MAU, NT‐proBNP, and hsCRP.

Patients and Methods

The design of i‐SEARCH Plus has been previously described. ⁴ The prospective registry was conducted according to local laws and regulations (§ 67 (6) Arzneimittelgesetz) and was duly notified to the federal authorities (Bundesinstitut für Arzneimittel und Medizinprodukte) and the federal panel doctors’ association (Kassenärztliche Bundesvereinigung). Ethics committee approval (Landesärztekammer Hannover) and patient informed consent was obtained.

Patients

Patients with hypertension (≥140 mm Hg or ≥90 mm Hg measured after 5 minutes of rest), being at least 18 years of age, and not having contraindications for treatment with irbesartan alone or in combination with 12.5‐mg hydrochlorothiazide (HCTZ) were included. Patients with impaired renal function, serum creatinine ≥2.0 mg/dL (177 μmol/L), urinary tract infection, febrile infection, menstruation, or pregnancy were excluded from participation because of the false‐positive microalbuminuria in these patients. Furthermore, patients with a history of drug or alcohol abuse were also excluded because of the expected low compliance.

Study Conduct

i‐SEARCH Plus was conducted on behalf of the steering committee by the Institute for Hypertension and Cardiovascular Research, Cloppenburg, Germany. Patients included in the registry had a complete clinical workup and were scheduled for further appointments at 3, 6, and 9 months, with a reduced number of clinical evaluations as compared with the baseline visit. After 12 months, patients were re‐examined and CV events were documented. ⁴ A total of 15% of office‐based cardiologists were monitored on site by comparing the case report forms with the respective source documents.

Risk Marker Definition and Determination

A dipstick test for microalbuminuria was provided to the participating physicians. Further urine and blood samples were obtained, shipped to the central laboratory at the St Josefs‐Hospital in Cloppenburg, Germany, and centrally analyzed. Blood samples were obtained only from Monday to Wednesday because NT‐proBNP has to be determined within 72 hours. The second morning urine collection was used for the determination of albuminuria.

At the central laboratory, MAU was determined with a nephelometric method. An albumin/creatinine ratio of ≥20 mg/g was considered to be microalbuminuric. The levels of hsCRP were determined by a commercial assay, as was the level of NT‐proBNP.

Statistical Analysis

Because of the exploratory character of the registry, the planned sample size of approximately 2000 patients was not based on a formal sample size calculation but resulted from feasibility deliberations. These were based on data published by Kistorp and colleagues, ⁵ who monitored the 3 above‐mentioned risk markers with respect to their CV prognosis for a period of 5 years in 537 patients without CV events at baseline.

The statistical analysis was performed descriptively and interpreted in an explorative way. Comparisons were made for a number of variables and analyzed using descriptive statistics. The number of patients was given for each variable and differences were calculated in patients, whose baseline and follow‐up values were documented according to the protocol. Patient subgroups were built for each single risk marker and separately evaluated. Differences in the number of CV events were calculated from logistic regression analysis (odds ratio [OR]) and adjusted for age and sex if appropriate. The analysis of data was performed with SAS 8.02 (SAS Institute, Cary, NC).

Results

Baseline Characteristics and Study Population

From May 2005 to August 2007, a total of 2173 patients were enrolled. Mean age of patients at baseline was 61.4±11.3 years and blood pressure (BP) was 159.8±20.1/93.4±11.9 mm Hg (Table I). A total of 46.9% received irbesartan alone, and 51.1% received irbesartan in fixed‐dose combination with 12.5‐mg HCTZ. Additional concomitant pharmacotherapy is displayed in Table II. This resulted in a mean BP of 137.6±17.8/83.0±10.3 mm Hg. At baseline median albumin/creatinine ratio was 9.90 (interquartile range [IQR], 5.76‐‐25.52) mg/g, hsCRP 2.46 (IQR, 1.16‐‐5.14] mg/L, and NT‐proBNP 89.28 (IQR, 38.63‐‐203.40) pg/mL (Table III). A total of 1649 patients completed the 12‐month follow‐up period (75.9% of the enrolled; 43.2% women).

Table I.

 Baseline Characteristics

demography	no. available/no. of total patients	mean± SD or%
Mean age, y	2173	61.4±11.3
Women, %	925/2173	42.6
BMI, kg/m2	2173	29.7±5.0
Waist circumference
Men, cm	1220/2145	107.3±12.2
Women, cm	900/2148	98.8±13.5
Systolic blood pressure, mm Hg	2173	159.8±20.1
Diastolic blood pressure, mm Hg	2173	93.4±11.9
Laboratory values
Total cholesterol, mg/dL	2161	216.64±42.97
HDL‐C, mg/dL	2161	63.69±17.05
LDL‐C, mg/dL	2052	118.36±35.21
Triglycerides, mg/dL	2161	181.02±123.50
HbA1c, %	2159	5.85±0.94
Serum creatinine, mg/dL	2161	1.06±0.27
Cardiovascular risk factors/concomitant disease
Previous MI/CAD family history	682/2160	31.4
No regular sports	1101/2156	50.7
Present smoker	264/2167	12.1
Previous smoker	628/2016	28.9
Family history of diabetes mellitus	589/2165	27.1
Diabetes mellitus	492/2162	22.6
Heart failure	103/2155	4.7
Atrial fibrillation	148/2162	6.8
CAD	370/2124	17.0
Ischemic stroke	55/2159	2.5
TIA	53/2162	2.4
PAD	76/2157	3.5
Carotid artery stenosis	60/2161	2.8

Abbreviations: BMI, body mass index; CAD, coronary artery disease; HbA_1c, glycated hemoglobin; HDL‐C, high‐density lipoprotein cholesterol; LDL‐C, low‐density lipoprotein cholesterol; MI, myocardial infarction; PAD, peripheral arterial disease; SD, standard deviation; TIA, transient ischemic attack.

Table II.

 Pharmacotherapy at Baseline

demography	no. available/no. of total patients	%at baseline
Irbesartan alone or in fixed‐dose combination
Irbesartan 75 mg	108/2173	5.0
Irbesartan 150 mg	709/2173	32.6
Irbesartan 300 mg	459/2173	21.1
Irbesartan 150 mg/  HCTZ 12.5 mg	298/2173	13.7
Irbesartan 300 mg/ HCTZ 12.5 mg	594/2173	27.3
Further antihypertensive pharmacotherapy
None	629/2173	28.9
Thiazides	220/2173	10.1
Loop diuretics	161/2173	7.4
α‐Blockers	69/2173	3.2
Calcium channel blockers	495/2173	22.8
β‐Blockers	1035/2173	47.7
ACE inhibitors	422/2173	19.4
Aldosterone antagonists	14/2173	0.6
Lipid‐lowering agents
None	1525/2173	70.2
Statins	553/2173	25.4
Fibrates	31/2173	1.4
Others	47/2173	2.2

Abbreviations: ACE, angiotensin‐converting enzyme; HCTZ, hydrochlorothiazide.

Table III.

 Risk Markers During Baseline and at Follow‐Up

biogenic risk markers	baseline mean± SD, median (IQR)	12 months mean± SD, median (IQR)	P value vs baseline (t test)
Albumin/creatinine ratio, mg/g	52.1±196.3, 9.90 (5.76–25.52)	49.0±315.7, 9.15 (5.13–20.92)	.7453
hsCRP, mg/L	4.6±8.7, 2.46 (1.16–5.14)	4.1±7.5, 2.01 (1.00–4.38)	.1031
NT‐proBNP, pg/mL	218.8±552.3, 89.28 (38.63–203.40)	217.0±620.3, 85.69 (34.93–187.00)	.9335

Abbreviations: hsCRP, high‐sensitivity C‐reactive protein; IQR, interquartile range; NT‐proBNP, N‐terminal pro‐brain natriuretic peptide; SD, standard deviation.

Risk Markers at Baseline in Relation to CV Events at 12 Months

During the 12 months of follow‐up, a total of 33 patients had any cardiac or cerebrovascular event compatible with the combined end point. Nine patients had newly diagnosed CAD, 1 patient had MI, 5 patients experienced stroke/TIA, and 5 patients died. Thirteen patients were hospitalized during follow‐up.

A positive test for MAU (≥20 mg/g) at baseline (n=791; 36.4%) identified patients with an increased risk for the combined end point of newly diagnosed CAD, MI, stroke/TIA, and death (OR, 3.06; 95% confidence interval [CI], 1.73–5.43; adjusted for age: OR, 2.67; 95% CI, 1.49–4.76) (Figure). MAU also increased the risk of 2 of its components: The OR for new CAD was 3.27 (95% CI, 1.39–7.68) in patients with MAU without any influence of age or sex. The OR for total death was 5.65 (95% CI, 1.76–18.07).

Figure.

 Forrest plot illustrating the impact of microalbuminuria on a composite end point of cardiovascular morbidity and mortality and its single components (adjusted for age). OR indicates odds ratio; CI, confidence interval; CAD, coronary artery disease; MI, myocardial infarction; TIA, transitory ischemic attack.

On the other hand, there was no influence of hsCRP or NT‐proBNP on the combined end point. A significant correlation of NT‐proBNP with total death (P<.0005) disappeared after adjusting for age and the presence of MAU (P<.189).

Correlations Among Cardiac Risk Markers

MAU‐positive patients had higher median values of both hsCRP (3.07 mg/mL in MAU‐positive vs 2.27 mg/mL in MAU negative patients) and NT‐proBNP (143.00 pg/mL vs 77.99 pg/mL). In further analyses there was a positive correlation between MAU and hsCRP (Spearman’s rho, 0.142; P<.01) and MAU and NT‐proBNP (Spearman’s rho, 0.275; P<.01).

Patients who developed MAU during the 12 months of follow‐up had higher median hsCRP and NT‐proBNP values than patients still MAU‐negative at follow‐up (P<.01; Table IV). Patients, who were MAU‐positive at baseline and stayed positive had higher median hsCRP and NT‐proBNP values than patients who became negative at follow‐up (P<.01).

Table IV.

 Changes of MAU Status During Follow‐Up and Median Values of hsCRP and NT‐proBNP^a

	patients mau negative at baseline and positive at 12 months (–/+)	patients mau negative at baseline and at 12 months (–/–)	patients mau positive at baseline and at 12 months (+/+)	patients mau negative at baseline and negative at 12 months (–/+)
hsCRP, mg/L	2.56	2.11	3.34	2.76
NT‐proBNP, pg/mL	105.50	73.64	150.60	99.77

Abbreviations: hsCRP, high‐sensitivity C‐reactive protein; MAU, microalbuminuria; NT‐proBNP, N‐terminal pro‐brain natriuretic peptide. ^a P value <.01 for all.

Correlations Between Risk Markers and Other Clinical Variables

There was a positive correlation between hsCRP and office heart rate (Spearman’s rho, 0.094; P<.001) and heart rate measured during ambulatory BP monitoring (Spearman’s rho, 0.070; P<.01), respectively. On the other hand, there was a negative correlation between NT‐proBNP and heart rate (Spearman’s rho, −0.153 for office and −0.277 for ambulatory BP monitoring, respectively; both P<.01). There was no such correlation of MAU with heart rate.

Correlations between cardiac risk markers and laboratory values such as hemoglobin A_1c, total cholesterol, high‐density lipoprotein and low‐density lipoprotein cholesterol, triglycerides, creatinine, and cystatin C were small but significant (P<.01). The best correlation was found between NT‐proBNP and cystatin C (Spearman’s rho, 0.398; P<.01).

Discussion

Laboratory risk markers such as MAU, hsCRP, and NT‐proBNP are used to refine the CV risk assessment provided by traditional risk factors, such as lipids and BP. We assessed the prognostic value of these risk markers in a hypertensive patient population in primary care treated with irbesartan either alone or in combination with HCTZ. We demonstrated that MAU at baseline but not hsCRP or NT‐proBNP was predictive for a combined primary end point of newly diagnosed CAD, MI, UAP, stroke/TIA, and death event in irbesartan‐treated patients.

Predictive Performance of MAU

It is well accepted that MAU, and even a lower level of urinary albumin excretion, is a marker of endothelial damage and indicative of an increased CV risk. ⁶ , ⁷ Of particular importance are the findings published by Hillege and colleagues ⁸ in a population‐based Dutch study, which identified MAU to be an independent predictor for CV events as well as for CV mortality (relative risk, 1.29; 95% CI, 1.18–1.40). ⁹ These findings are well supported by the present analysis in which MAU was predictive of subsequent CV events, and adjustment of other risk markers for the presence of MAU demonstrated its relative importance.

These results, however, have to be interpreted with respect to the influence of the irbesartan treatment introduced at the baseline visit. Irbesartan has been shown to influence the degree of albuminuria in patients with type 2 diabetes and hypertension. ² , ¹⁰ It was surprising, however, to see in this study that irbesartan treatment did not change the relative excretion of albumin into the urine. There are different reasons that may have resulted in this finding. First, the level of albumin excretion was low compared with randomized controlled trials, and studies such as the Diabetic Retinopathy Candesartan Trials (DIRECT) with candesartan ¹¹ and the recent Randomized Olmesartan and Diabetes Microalbuminuria Prevention (ROADMAP) study with olmesartan have resulted in a failure of these drugs to lower low‐grade albuminuria. Second, the uncontrolled design of i‐SEARCH Plus does not enable us to compare the results with a control group, which is unfortunate but cannot be corrected. On the other hand, an observational study by Lehnert with irbesartan has also shown a reduction of microalbuminuria (based on a dipstick test). ¹² Third, the mean doses have been lower compared with those shown to be effective in randomized clinical trials. Therefore, the findings might not show the true correlation of MAU at baseline with CV events in a hypertensive primary care population but illustrates that despite irbesartan treatment, there is a residual CV risk that can be recognized by urinary albumin level determination. ¹³

The determination of MAU was valuable up and above the two other risk markers under investigation (hsCRP and NT‐proBNP) in our study. This essentially confirms a recent study by Olsen and colleagues ¹⁴ who reported that additionally determining MAU but also hsCRP revealed more prognostic information than using the HeartScore alone in a low‐risk population, with hypertension being no inclusion criterion.

Predictive Performance of hsCRP

hsCRP is an inflammatory marker for early atherosclerosis and has been associated with an increased risk for CV events. ¹⁵ , ¹⁶ Nozaki and colleagues ¹⁷ identified hsCRP as a significant and independent predictor of future CV events (hazard ratio, 1.468; 95% CI, 1.150–1.875). Furthermore, large prospective trials (Women’s Health Study, Physician’s Health Study) have shown that slightly increased hsCRP values at baseline in healthy patients are associated with a 2‐fold increase in risk for MI, stroke, and the development of diabetes. ¹⁵

In contrast, the present study found no significant role of hsCRP for the prediction of CV events in a hypertensive patient population in primary care. hsCRP was only slightly correlated with traditional risk factors and heart rate. These results, however, are compatible with findings from Olsen and associates, ¹⁴ who found that the determination of hsCRP did not add new prognostic information in patients with low to moderate CV risk. They suggested that elevated hsCRP levels simply reflect slightly elevated traditional risk factors in younger patients. This might also be a likely explanation for the findings in our analysis. Although our patients were older and had a certain burden of prior CV disease, they were well treated and controlled with respect to a number of risk factors, which is reflected by the low absolute incidence of CV events during 1 year. Another issue to be considered when assessing the role of hsCRP is the initiation of ARB treatment at baseline. Irbesartan (but not ACE inhibitors) has been shown to reduce hsCRP in patients with CAD and arterial hypertension. ¹⁸ Both factors may explain the failure to assign a predictive role of hsCRP for CV events in this dataset.

Predictive Performance of NT‐proBNP

The prognostic value of NT‐proBNP has mainly been established in patients with heart failure/volume overload, ¹⁹ , ²⁰ acute coronary syndrome, ²¹ and especially LVH. ²² , ²³ In a study by Kistorp and colleagues, ⁵ risk stratification for NT‐proBNP, hsCRP, and MAU was carried out in a general Danish population (764 patients, 50–89 years of age, without heart and renal failure). In that study, NT‐proBNP was a better predictor than both CRP and urinary albumin together with respect to overall outcomes. Also, Wang and coworkers ²⁴ proclaimed brain natriuretic peptide to be the most (out of 10) informative biomarker for predicting major CV events in a routine examination in the Framingham Heart Study in 3209 patients. In addition, Andersen and colleagues ²⁵ found incremental NT‐proBNP values in normoalbuminuric diabetic patients with left ventricular hypertrophy.

We found that NT‐proBNP was correlated to traditional risk factors and had a positive correlation to systolic and a negative correlation to diastolic BP. This is in complete agreement with previous studies; however, its missing correlation with CV events is in disagreement with previous reports assigning a more important role to NT‐proBNP. ²⁶ Overall it appears plausible to consider NT‐proBNP a valuable risk marker in a defined patient population ¹⁹ , ²¹ but not in hypertensive patients in general.

Limitations

i‐Search Plus was a nonrandomized, open study in patients with hypertension who were given irbesartan treatment at baseline. Therefore, the results cannot be generalized to all patients treated for hypertension, but are of particular value because of the standardized RAS‐blocking antihypertensive treatment regimen (all patients received irbesartan). Further unknown variables influencing the associations could not be considered for in randomized controlled trials. The study duration was limited to 12 months, which is in contrast to a number of epidemiologic studies that followed their patients for up to 5 years. Associations detected in these long‐term studies might not have become apparent in our study because of the short follow‐up and low event rate.

Conclusions

Microalbuminuria is predictive of future CV events in a hypertensive patient population recruited at office‐based cardiologists despite evidence‐based angiotensin receptor blocker treatment. These findings confirm previous reports on the prognostic value of MAU, establish its incremental value over hsCRP and NT‐proBNP, and underline its particular value in predicting CV risk.

Acknowledgments and disclosures:  We would like to thank the participating physicians, their assistants, and all patients observed during the registry. The study was supported by an unrestricted educational research grant by Sanofi‐Aventis Deutschland GmbH to Dr Ulrich Tebbe, Detmold. UT, PB, SL, FdH, RS, MB, and JS declare receiving either research support, speakers’ honoraria, or consulting fees from Sanofi‐Aventis and other pharmaceutical companies. The study has been funded by a research grant from Sanofi‐Aventis. WDP is an employee of Sanofi Aventis Deutschland GmbH.

Author’s contributions:  All authors were involved in the design of this registry. UT, SL, and JS supervised the conduct and the acquisition of data. PB drafted the manuscript and all authors revised the manuscript for important intellectual content. All authors read and approved the final manuscript.