Background: The use of thiazide diuretics in the treatment of hypertension (HT) is widely considered a first line treatment, given the efficacy and low cost of this class of drugs. This indication is not unanimous, because thiazides can cause metabolic alterations and other side effects increasing cardiac and cerebrovascular risk, which reduce compliance to treatment and increase health care system cost. However, large intervention trials in HT suggest that the improvement in cardiovascular prognosis of HT patients depends more on follow-up procedures than on type of drug used. Furthermore, the investigators have documented improved compliance to antihypertensive therapy by implementing cooperation between general practitioners (GPs) and HT specialists.

Objectives: In a multicenter, open label randomized study the investigators will compare the persistence on therapy of thiazides versus other treatments, as a first line antihypertensive therapy, in a clinical setting characterized by a strict cooperation between GPs and HT specialist. The investigators will also analyse candidate genes with impact on drug-induced metabolic alterations to elucidate the pathophysiology of these phenomena.

Methods: 260 GPs will recruit 2600 hypertensive patients with indication to pharmacological treatment and randomise them to starting treatment with chlortalidone (12.5 to 25 mg daily, 1300 pts) or a GP decided single drug (excluding thiazides) or combination therapy at highest tolerated dose. In both groups any other class of antihypertensive drugs can be added over time in order to achieve blood pressure control (<140/90 mmHg). Follow-up will last 2 years. Blood sample and urine analyses, carotid and cardiac ultrasound will be performed at baseline and scheduled time points. Genotyping will be performed by sequencing. Data will be collected and stored using a web based centralized Case Report Form (CRF) Expected results: Results will highlight whether a follow-up strategy based on tight cooperation between GPs and HT specialists can allow the use of thiazides as first line antihypertensive therapy without any negative effect on persistence, adherence and efficacy of the treatment. These data can be used to reduce total burden of the Health Care System in HT by replacing more expensive drugs with diuretics in the 50% of hypertensive patients, who do not receive this class of drugs. Furthermore, the pharmacogenetic approach may clarify the pathophysiological mechanisms of drug-induced metabolic side effects

Background and rationale

Many comments have been issued about similarities and differences between 2003 American and European guidelines for the management of arterial hypertension (1,2), especially after the publication of the Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial (ALLHAT) study (3-5). Both guidelines agree that the majority of hypertensive patients require more than one medication to achieve optimal blood pressure control and, consistent with the most recent findings, the need of expensive, large clinical trials to demonstrate the superiority of one medication on another is probably over (3,6). What is clear now is that blood pressure must be optimally controlled to reduce risk of cardiovascular mortality and morbidity in population, no matter which combination of medications is used. Comparison between single medications is substantially academic and conceals the reality: in all trials, in most circumstances comparison is between combinations, more than single medications. In the Losartan Intervention for Endpoint Reduction Study (LIFE study) (6), for instance, there has been a comparison between the combination of losartan with low dose hydrochlorothiazide versus the combination of atenolol with the same diuretic.

The difference between the seventh Joint National Committee (USA) on Prevention, Diagnosis and Management of Hypertension (JNC VII) and European Society of Cardiology (ESC) and European Society of Hypertension (ESH) 2003 guidelines, often presented as substantial, is in the priority that Americans still give to the diuretic therapy, consistent with the most waited results of the ALLHAT (3).

The ALLHAT was a randomized, double-blind, multicenter clinical institutional trial, entirely sponsored by the National Heart, Lung, and Blood Institute (NHLBI) (3), and designed to determine whether occurrence of fatal or nonfatal coronary heart disease is lower for high risk hypertensive patients treated with amlodipine, lisinopril, doxazosin, or chlorthalidone. The protocol was also approved by an independent Review Committee external to the NHLBI (7). The ALLHAT recruited 9,000-15,000 participants/intervention arm (total: 33,357), and the follow-up was quite long (4-8 years). The doxazosin arm was closed prematurely because of higher mortality (8). Despite a number of important limitations of this study (4,9), the overall impact of its findings remains very high.

The baseline characteristics of the ALLHAT participants were substantially similar in the 3 arms completing the study. Chlortalidone was significantly more effective than both amlodipine and lisinopril in achieving optimal control of blood pressure at year 1 and year 2 and more effective than lisinopril in blood pressure control at year 3, 4 and 5 (all p<0.001). Consistent with the better control of blood pressure, chlortalidone tended to provide a 10% more protection for combined fatal and non fatal coronary heart disease than lisinopril whereas protection was similar for amlodipine, though this protection was less evident than the superiority in controlling blood pressure. In fact, the difference in risk profile versus the lisinopril arm achieved statistical significance only for elderly and African-American participants.

Another relevant finding from the ALLHAT concerns the incidence of congestive heart failure that was substantially less with chlortalidone than with either amlodipine or lisinopril.

There are many recurrent criticisms to all trials showing the superiority of diuretics over other medications in controlling blood pressure. First of all, worsening of glucose metabolism, due to thiazide effects, is matter of concern (4,14-18). Also the ALLHAT authors report a higher incidence of diabetes in the diuretic group which, however, did not affect the outcome results (3). However, this concern is also sustained by recent evidence of the dangerous effect of incident diabetes in patients with arterial hypertension (15).

One factor that might aggravate glucose metabolism during therapy with diuretics is the underestimated effect of hypokalemia, which interferes with glucose-stimulated insulin release, a condition that might be aggravated by association with beta-adrenergic block (19,20). Among other beneficial effects, including improvement in blood pressure control, correction of hypokalemia prevents or significantly reduces thiazide-induced hyperglycemia (21-24). It is possible that low-dose potassium-sparing diuretics by preventing hypokaleemia might also prevent metabolic effects of thiazide, at the same time enhancing the antihypertensive effects of thiazide (25), but, this hypothesis has never been tested in ad hoc trials.

The second concern about the use of diuretics is the risk of low therapeutic compliance of hypertensive patients, due to the diuretic effect and other hypokalemia-related side-effects (26,27). These few reports, however, contrast with the evidence that quality of life is improved by long term therapy with diuretics (28-30). The effort, therefore, should probably be direct toward persistence of initial treatment with diuretics, a goal that might be reached by improving the procedures of follow-up, for instance, as we have recently proposed, by implementing internet-based digital networks, connecting hypertension specialists with general practitioners (31). The Campania Salute (CS)network is a system that was set up in 1995 by us (31). It is an italian regional network system aimed at improving the management of essential hypertension by integrating the activity of general practitioners (GPs) with hypertension specialists. This network includes about 12.000 hypertensive patients followed by 23 outpatient hypertensive clinics allocated in different Community Hospitals in the Campania Region and 60 GPs, homogeneously allocated in the regional area, referring to the Hypertension Clinic of Federico II University Hospital in Naples (coordinating centre). GPs were randomly selected among a pool of physicians referring their patients to the Hypertension Clinic at Federico II University. Through the CS system clinical data detected at each visit can be shared between the coordinating center and the peripheral units.Low-risk hypertensive patients continue their follow up in the peripheral units, whereas high-risk hypertensive patients are more closely followed up by the coordinating centre, which also evaluates target organ damage and associated diseases. Patients' information is shared through on-line access to the remote database, integrated by smartcards. The smartcard belongs to patients and contains his/her personal identification number (PIN). This PIN allows users to access the file of the patient stored in the database. Each physician has also his own PIN to access into the database, limited to his/her own patient files. Clinical data are upgraded at each visit by GPs and physicians of both peripheral centres and coordinator centre. Access to the remote database allows users to read all clinical and laboratory data, as well as tracking electrocardiography (ECG) and cardiac and vascular ultrasound images. In addition, the smartcard also works as portable database in which identification and clinical data are reported. By virtue of a central database, data of individual patients can be stored, updated and retrieved directly on-line by participants in the project. The restricted access to individual data requires the pre-assigned identification code of both the patient and the relevant remote units. The central database uses Wincare software (TSD Projects, Milan, Italy) which contains separate sheets for medical history, physical examination, biochemistry electrocardiography, cardiovascular ultrasound, other imaging tests and ambulatory blood pressure monitoring. The last update of an individual patient's record can also be downloaded and stored in the patient's personal smartcard. We started the CS project with the aim of obtaining a stronger interaction between GPs and hypertension clinics, by providing a direct link and accessible patients' records. Blood pressure control was improved with our referral system, since better overall results were obtained if the patient was followed within the network. Indeed, at the end of the observation period, 51% of patients in the CS group had a blood pressure below 140/90 mmHg, a percentage comparable to that of patients included in clinical trials.

This follow-up strategy also allowed an active pharmacovigilance procedure, with side effects promptly reported to the GPs, which prevented the occurrence of difference in compliance among the various antihypertensive treatments. In addition, this kind of follow-up allowed by means of pharmacogenetic studies to demonstrate that the occurrence of side effects may be predicted from individual genotype (32). Indeed, we have recently reported that in patients bearing β2AR gene Glu27 variant or the β3AR gene Arg64 variant there was a larger occurrence of hypertriglyceridemia, alone or in combination with elevated cholesterol levels. Furthermore, the β2AR Glu27variant significantly associates with hypetriglyceridemia in a cumulative fashion. The risk of developing this side effect after β-blockade was four-fold higher in patients homozygous for the β2AR Glu27 variant than in β2AR27Gln allele. This result not only allowed the identification of patients at high risk to develop metabolic complications to chronic β-blockade treatment, but also contributed to elucidate the pathophysiological mechanisms which mediate these side effects, raising the possibility to prevent them.

Objectives of the study

This study has been designed to assess whether a follow-up strategy based on a strict cooperation between GPs and hypertension specialists allows the use of diuretics as first-line antihypertensive treatment with a persistence on assigned therapy equivalent to that achieved by using any other first line antihypertensive therapy. Assessment of safety and efficacy for controlling cardiovascular risk will be also performed as secondary endpoint. In fact, in contrast with trials comparing single-drug effects, this study compares two strategies of antihypertensive management, based on either real-word prescriptions or a regimen in which thiazide diuretics represent a forced first line antihypertensive therapy. If our hypothesis will be demonstrated diuretics might be suggested as an efficient and economic first line antihypertensive treatment, on which build up optimal antihypertensive therapy by adding other class drugs, in all patients, provided that the follow-up procedure is based on the proposed organization. This approach will be of great utility for the National Health Care System to reduce costs, since, a large part of the economic burden is related to the use of antihypertensive medications more expensive than thiazides, as first line agents, in particular so far only 40% of the hypertensive patients receive diuretics in their therapy.

Finally, the pharmacogenetic study is focused on characterization of polymorphisms of candidate genes associated to development of metabolic side effects of diuretics, to help understanding of mechanisms underlying these adverse events. This kind of information will help to prevent the occurrence of adverse events by the use of adeguate combination treatment, thus resulting in the further reduction of the cost of antihypertensive treatment due to the reduction of the number of patients that discontinue therapy for occurrence of adverse events.

Study design

This is a multicenter, open label, randomized study to compare the effects of an antihypertensive strategy using a thiazide diuretic as first-line, versus the use as initial therapy of other antihypertensive treatments. All the analyses will be performed by personel blinded to treatment. The study will be performed in collaboration with the Società Italiana Medicina Generale (SIMG), Sezione Campania, and the Società Italiana Ipertensione Arteriosa (SIIA), Sezione Campania.

Study population. The recruitment phase will last 8 months. During this period 2600 patients will be enrolled, in the offices of 260 GPs' with documented previous experience in controlled studies, performed according to recommendations of Good Clinical Practice, and availability to access to Internet. Selected GPs will be trained to the use of the web-based database on which they will store the required information of patients participating into the study. This training period will last a week and will be supervised by the coordinator's center. Exemplificative print outs of the web-based CRF are available for evaluation on the web site www.campaniasalute.it.

GPs are required: 1) to record a full medical history, including smoking and drinking habits, based on a pre-defined clinical record; 2) to collect demographic and anthropometric measures (height, weight, waist circumference at the iliac crest); 3) to perform a complete physical exam. At baseline and at each visit thereafter, seated office blood pressure will be measured in triplicate using a manual sphygmomanometer according to international guidelines. Measurements will be rounded to the closest 2 mmHg interval.

Inclusion criteria: Hypertensive patients will be 18 to 75-year old. Eligible patients are required to have stage Ic or II essential hypertension, and to be previously untreated or poorly controlled. They will be selected by GPs participating into the study. Similar to untreated patients, those with poor control of blood pressure under multiple-drug therapy will start treatment with one single drug, which will be titrated to the highest dose before adding subsequent medications, based on the GP's judgement.

Hypertension will be defined according to 2003 ESH/ESC guidelines (1). Blood and urine tests will be performed, according to guidelines for Hypertension Management For General Practitioners (GP) of the Regione Campania (see Bollettino Ufficiale Regione Campania, number 11, 18/02/2002, allegato A). This screening includes cell blood counts (CBCs), serum creatinine, sodium, potassium, uric acid, total cholesterol, triglycerides, HDL-cholesterol, glucose, urine analysis and EKG. LDL will be calculated starting from the total cholesterol, triglyceride and HDL-cholesterol.

Exclusion criteria. Women in fertile age not using recognized contraceptive methods, or pregnant or nursing will be excluded from the protocol, since the use of many antihypertensive drugs is contraindicated in pregnancy and lactation. Patients will be excluded when presenting with documented coronary or cerebrovascular events in the previous 6 months, NYHA class higher than 1, history of congestive heart failure, secondary hypertension, cancer disease, renal disease (serum creatinine >2 mg/dl), liver cirrhosis or severe dysfunction, or any other health problem that may interfere with the projected 2 year follow-up. Data will be stored in an electronic database located in the Coordinating Centre, to which GPs may have access for uploading data on a daily base, using personal, encrypted, login and password. Eligible patients will be asked for written informed consent and thereafter referred to the identified Hypertension Specialist Centre located in their areas, for end-organ damage evaluation by echocardiography, carotid ultrasound and urine dip-stick. These data will be stored in the central database. After local echocardiographic evaluation, patients showing left ventricular Ejection Fraction < 45% will be excluded from the study.

Eligible patients will be asked for blood sampling for genomic DNA analysis and then randomised by the coordinating centre to either diuretics or other treatment. This latter will be decided by the GPs. Randomization will be organized in permuted blocks of 10 patients for each GP, half of which will be assigned to diuretics. The randomization code will be communicated to the referring GP by e-mail.

Blood samples for genetic analysis and signed informed consents will be sent to the coordinating centre for storage. Blood samples will be coded and anonymously processed for genetic analysis by the Department of Pharmacology of FEDERICO II University of Naples. The data resulting from this analysis will be stored in the patient CRF page.

Intervention: Drugs will be administered orally. GPs should use chlortalidone (12.5-25 mg daily) in the arm with compulsive thiazide diuretic as first line. In the alternative arm, GPs may choose any appropriate single-drug (excluding thiazide diuretic) or combination therapy, as first-line, at the tolerated dose.

After randomization, patients will be evaluated monthly at the GPs' office, for therapy adjustment, to achieve blood pressure normalization (i.e Systolic Blood pressure >140 mmHg, and Diastolic Blood pressure >90 mmHg). In the thiazide arm, if blood pressure normalization is not achieved not even with the maximal dose, it will be possible to add any other classes of antihypertensive drugs. Once blood pressure normalization is achieved, GPs will monitor blood pressure once every 2 months, at the renewal of the drug prescription. Blood pressure values will be stored in the central database. At each visit, GPs will record drug therapy, including concomitant medications, evaluate the compliance to assigned antihypertensive regimen, and monitor and record adverse events by reporting all data in the CRF. After two years from randomization, patients will be checked for blood and urine tests and referred to the Hypertension Specialist Centre for echocardiography, carotid ultrasounds and urine dip-stick. This data will be stored in the central database.

Information retrieval: At each visit, the GPs will record drug therapy, including concomitant medications, will evaluate the compliance to the assigned antihypertensive medications by pill count, and will monitor and record adverse events by reporting all data in the CRF.

Pre clinical cardiovascular disease will be assessed by echocardiography and carotid ultrasonography. All ultrasound exams will be sent to the Reading Center at Federico II University Hospital and will be processed, according to procedures described in the annex.

Monitoring of the study: All data will be reported on a specifically designed electronic clinical research form (CRF) (see web site: www.campaniasalute.com), and will be transferred to the Coordinating Center for data storage and analysis. The Steering Committee will appoint a Data Coordinating Committee to evaluate all CRFs on a continuous way to ensure quality and objectivity of the analysis performed by trained professionals. In order to monitor for patient security, GPs will be asked to actively monitor periodically for adverse events, by asking patients at the time of drug prescription renewal for the occurrence of symptoms or signs that can be related to aggravation of their condition or adverse events of therapy. In selected cases, GPs can refer the patients to the Hypertension Specialist Centre for eventual instrumental and or blood and urinary analyses. Adverse events will be reported in the digital CRF.

Sample size estimate: The main outcome to be tested is whether persitance on therapy of an antihypertensive regimen based on diuretics as first choice is equivalent to that obtained in a free regimen using any other antihypertensive medication as first choice (equivalence study).

As secondary outcomes, reduction of left ventricle (LV) mass and carotid intima-media thickness will be evaluated as markers of preclinical cardiovascular disease, under the hypothesis that improvement of end-organ damage under diuretic-based treatment will not be different from the treatment based on other antihypertensive medications (equivalence). The reduction of the ESH/ESC risk-score will also be evaluated, under the same equivalence hypothesis. Sample size was primarily estimated for the primary outcome, but afterward tested on power also on secondary outcomes. See Annex for details

Organizational characteristics: The study will be governed by a Steering Committee chaired by the Principal Investigator (Prof. Bruno Trimarco). The study will be performed as a collaborative effort of 260 general practitioners (with previous experience in scientific initiatives), specialist centers (Community Hospitals and University Hospitals in which specialized evaluation of hypertension related organ damage is routinely evaluated). The Department of Medicina Clinica, Scienze Cardiovascolari ed Immunologiche, division of Coronary Intensive Care Unit and High Blood Pressure Center will act as a Coordinating Center, run by Professor Trimarco, PI. The central database will be stored at this center, and the other specialist centers and GPs will have remote access to it through encrypted login and password. The web-based access to the database has already been implemented at the Coordinating Center, within the Campania Salute Project, a regional network of Community Hospitals and GPs.

Echocardiograms and carotid ultrasounds will be performed at the Federico II University Hospital or in peripheral centers, under a standardized protocol, which will be distributed in an electronic format (CD-ROM). All studies will be directly transmitted through the Internet to the Reading Center.

Feasibility: The Principal Investigator, Bruno Trimarco, has a long and extensive experience in running trials of antihypertensive treatment and management of hypertension (see CV). As Director of the Coordinating Institution he will personally assure full support of the institution facilities, and of professional help in monitoring and statistics.

The Ultrasound Reading Center has wide experience in centralized reading of studies on LV hypertrophy and function as well as in studies on arterial structure and has been involved in a number of international multicenter trials (35,40). The Reading Center is provided of 4 work stations for echocardiographic reading and 2 for carotid ultrasound, with high level of security for preservation of data and privacy. All echocardiograms will be classified with a reception number which will join the recruitment number of the participant (every participants will have 2 identification numbers in addition to the number of identification document).

Timing: The study will last 3 years. The first eigth months will be spent for recruitment and randomization. The follow-up will last 2 years. Analysis of data will be performed ad interim, as soon as the last recruited patient completes the intermediate evaluation after one year of follow-up. Final main analysis will be performed right after the conclusion of follow-up of the last recruited patient. A number of analysis concerning secondary end-points and including every ancillary study that might be proposed from the Steering Committee or the participating Hypertension Specialist Centres, will be implemented thereafter.

Ethical aspects. We have tried to minimize possible therapy related side effects to those that are usually observed in the practical clinic. Indeed, all treatments and dosages are those that are usually adopted by general practitioner for the daily practice. Therefore, we do not expect any additional risk for patients that are enrolled in the study. The complications that are associated to thiazide treatment will be prevented by the use of maximal doses that are in the low range of therapeutic effect, and close to the regimen that currently used in daily practice. As for intromission in the private sphere of the patients, the data will be nominally entered in the database by the physician using a login/password protected web-based 32bit encrypted connection, and available only for this research purposes after given informed consent by the patients.

Inclusion Criteria:

• Hypertensive patients will be 18 to 75-year old.
• Eligible patients will be asked for written informed consent and thereafter referred to the identified Hypertension Specialist Centre located in their areas, for end-organ damage evaluation by echocardiography, carotid ultrasound and urine dip-stick.
• Eligible patients are required to have stage Ic or II essential hypertension, and to be previously untreated or poorly controlled. They will be selected by GPs participating into the study. Similar to untreated patients, those with poor control of blood pressure under multiple-drug therapy will start treatment with one single drug, which will be titrated to the highest dose before adding subsequent medications, based on the GP's judgement.
• Hypertension will be defined according to 2003 ESH/ESC guidelines (1). Blood and urine tests will be performed, according to guidelines for Hypertension Management For General Practitioners (GP) of the Regione Campania (see BURC number 11, 18/02/2002). This screening includes cell blood counts (CBCs), serum creatinine, sodium, potassium, uric acid, total cholesterol, triglycerides, HDL-cholesterol, glucose, urine analysis and EKG. LDL will be calculated starting from the total cholesterol, triglyceride and HDL-cholesterol.

Exclusion Criteria:

• Omen in fertile age not using recognized contraceptive methods, or pregnant or nursing will be excluded from the protocol, since the use of many antihypertensive drugs is contraindicated in pregnancy and lactation.
• Patients will be excluded when presenting with documented coronary or cerebrovascular events in the previous 6 months, NYHA class higher than 1
• History of congestive heart failure
• Secondary hypertension
• Cancer disease
• Renal disease (serum creatinine >2 mg/dl)
• Liver cirrhosis or severe dysfunction, or any other health problem that may interfere with the projected 2 year follow-up.

Data will be stored in an electronic database located in the Coordinating Centre, to which GPs may have access for uploading data on a daily base, using personal, encrypted, login and password. These data will be stored in the central database. After local echocardiographic evaluation, patients showing left ventricular Ejection Fraction < 45% will be excluded from the study.