Of the estimated 50 million hypertensive patients in the United States, only one third have their blood pressure (BP) controlled to a goal of 140/90 mm Hg.1 To address this serious problem, federally sanctioned treatment guidelines are rewritten every few years with the intention of improving the care of hypertensive patients. In 2003, the Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure (JNC 7) was released with the stated purpose of providing “an evidence based approach to prevention and management of hypertension.”1 One of the major conclusions of this report was “More than two-thirds of hypertensive individuals cannot be controlled on one drug and will require two or more antihypertensive agents selected from different drug classes.” This conclusion has been challenged because of the limited design and interpretation of the randomized controlled treatment trials cited by JNC 7. Those trials used a parallel design in which two or more groups of patients were each assigned to a different Step 1 drug.2,,4 This experimental approach also has been criticized conceptually because it assumes that the same pathophysiologic mechanism causes hypertension in all patients and that a diuretic should therefore be a universally effective, first line treatment for all.5–10 This monolithic treatment strategy belies the evidence that a spectrum of mechanisms contributes to the pathophysiology of hypertension and that patients can be stratified accordingly.6,11,12 Those with predominantly sodium volume–dependent (“V”-type) hypertension can be identified by a plasma renin activity (PRA) level below 0.65 ng/mL/h (<5 U/mL). At the other end of the spectrum are patients with renin-dependent (“R”-type) hypertension, characterized by a PRA level at or above 0.65 ng/mL/h (≥5 U/mL). This classification has been shown to simplify the selection of effective primary antihypertensive drug treatment as monotherapy.11,13–15 Thus, V hypertension is most effectively treated with natriuretic drugs (ie, diuretics, calcium channel blockers, -adrenergic blockers), whereas R hypertension is most responsive to drugs that block renin-angiotensin system activity (ie, angiotensin converting enzyme [ACE] inhibitors, angiotensin II receptor blockers, -blockers, central 2-adrenergic agonists [methyldopa, clonidine, reserpine]). In support of this pathophysiologic heterogeneity, patients who have a favorable antihypertensive response to monotherapy with an R drug do not usually respond to monotherapy with a V drug. Conversely, R drugs are often ineffective in V hypertension. Failure to recognize these patterns means that half or more of patients could be given drugs for prolonged periods that have little chance of providing benefit. Moreover, the incidence of side effects and non-compliance increase when multiple drugs are prescribed. The concept that selection of effective antihypertensive treatment is related to the underlying pathophysiologic mechanism has rarely entered the arena of the clinical drug trial. However, antihypertensive treatment has been optimized in crossover rotation studies of four major drug classes.14,15 In those studies, every hypertensive patient was treated with monotherapy for 1 month with a drug from each of four drug classes: an ACE inhibitor or -blocker from the R drug group and a calcium channel blocker or diuretic from the V drug group. Each drug period was followed by a washout period until all four drugs were tested.14 A goal pressure of below 140/90 mm Hg was reached in 73% of patients during monotherapy, although fewer than 40% of patients reached that target with the first drug in the sequence. When a more rigorous goal pressure (≤135/85 mm Hg) was sought, slightly less than one half of all patients reached this target with monotherapy, but more than 90% of all study patients achieved that goal when a single R drug was paired with a V drug. The pretreatment PRA level was a significant predictor of BP response to therapy with an ACE inhibitor—the higher the pretreatment PRA, the greater the magnitude of the fall in pressure during treatment. Moreover, drug efficacy was also highly correlated within R and V drug classes, so that achieving goal pressure during monotherapy with an ACE inhibitor predicted a similar response to monotherapy with a -blocker. There was also a direct correlation between diuretic and calcium channel blocker BP responses. However, those who failed to respond to an R-type drug were most likely to respond to a V-type drug, and vice versa. By contrast with this mechanistic approach toward effective antihypertensive drug selection, the JNC 7 treatment algorithm instead stratifies patients according to whether they have co-morbid conditions such as diabetes, renal insufficiency, or heart failure.1 For those without such indications, JNC 7 recommends primary treatment with a thiazide-type diuretic for Stage 1 hypertension. For those with Stage 2 hypertension, first line therapy is a two-drug combination, including a thiazide-type diuretic and a member of “some other drug class.” If BP is not controlled at increased doses of these agents, then more are added without subtracting the ineffective drugs. What is the evidence to support these JNC 7 treatment guidelines? The drug trials cited by JNC 7 used a parallel design2–4 in which two or more groups of patients were each assigned to a different Step 1 drug. If the BP goal was not achieved, then one or more drugs were added stepwise while the previous medication was continued. Substitution or discontinuation of ineffective drugs was not an option in those protocols. This stepped care approach ignores the different pathophysiologic mechanisms that can cause hypertension in different patients. Moreover, although effective monotherapy can be achieved in a majority of patients, the first drug chosen is often not the most effective drug.14 The Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial (ALLHAT) is the largest antihypertensive drug study ever conducted.16 More than 40,000 hypertensive patients enrolled in ALLHAT were treated during a 5-year follow-up. It used a parallel design (ie, eligible patients were randomized to one of the treatments to be compared), providing the cornerstone for the stepped-care approach endorsed by JNC 7. The study population was at high risk for cardiovascular disease and stroke because of its age (mean age 67 years) and relatively high prevalence of African Americans, diabetic individuals, and patients with preexisting cardiovascular disease. Patients were randomized to receive a thiazide-type diuretic (chlorthalidone), calcium channel blocker (amlodipine), ACE inhibitor (lisinopril), or -blocker (doxazosin). The doxazosin limb was discontinued because of a reportedly higher incidence of heart failure compared with the chlorthalidone group. If a target BP below 140/90 was not reached with up-titration of the Step 1 drug, then a Step 2 drug was added, and if necessary, a Step 3 drug as well. The Step 2 drugs (ie, atenolol, clonidine, reserpine) used in ALLHAT are R-type drugs and, thus, are most effective as monotherapy in R-type hypertension and are less effective in V-type hypertension.17 With this selection of medication, more than 60% of the patients required that a Step 2 or Step 3 drug be added by year 5 of the study. Based on the results of studies described above and on others, one could reasonably predict that, in ALLHAT, when the Step 1 R-type drug, lisinopril, was ineffective, then V-type hypertension was probably present.13–15 In that case, addition of a Step 2 drug (also R-type drugs) would be unlikely to lower BP further. Thus, limiting the treatment options in the lisinopril limb solely to R-type drugs most likely accounted for the finding that twice as many in the lisinopril group as in the chlorthalidone group required progression to Step 3 treatment. Conversely, patients in the V-type Step 1 groups (ie, chlorthalidone or amlodipine) with persistent hypertension probably had an R-type mechanism; therefore, addition of a Step 2 R-type drug was likely to improve their BP without requiring progression to Step 3 treatment. Altogether, these observations suggest that the protocol design of ALLHAT placed the lisinopril treatment group at a disadvantage by comparison with the chlorthalidone group. This apparent design bias in ALLHAT has important implications because the lisinopril group was less likely than the chlorthalidone group to achieve the goal BP at 5 years of treatment. Although the difference in mean systolic BP between these two groups was relatively small (2 mm Hg higher in the lisinopril group), its pathophysiologic significance is indicated by the 15% higher risk of stroke and 10% higher risk of cardiovascular complications in the lisinopril group. Further analysis showed that these treatment group differences in BP and risk of cardiovascular disease were disproportionately high in African-American patients and that the increased risk of stroke occurred solely in that subgroup. Specifically, in African-American patients, mean systolic BP was 4 mm Hg higher in the lisinopril group than in the chlorthalidone group. This finding was associated with a 40% higher risk of stroke and 19% higher risk of cardiovascular disease—approximately twice that of the non-African-American patients. Was there an identifiable feature in the design of ALLHAT that could account for the racial differences in BP and risk of hypertensive complications? Alderman et al have reported that African-American hypertensive patients are more likely than non-African-American patients to have low renin level.18 In view of the redundant use in the lisinopril group of R drugs as both Step 1 and Step 2 treatment, it is likely that African-American patients in that group would have been less likely to have their BP controlled and, thus, were at higher risk for hypertensive complications than the chlorthalidone group. The possibility that interactions between these racial, pathophysiologic, and drug class variables may have confounded the initial interpretation of ALLHAT is supported by a recent reanalysis of the ALLHAT data by Cutler et al.19 This subsequent analysis was limited to subgroups of patients in whom the goal BP was achieved by monotherapy with either lisinopril or chlorthalidone and thus was not confounded by the selection of Step 2 drugs. Compared with the chlorthalidone monotherapy group, fewer patients in the lisinopril monotherapy group were African American, and there were no BP differences between these drug groups during treatment. Moreover, the chlorthalidone and lisinopril groups had equivalent relative risks of coronary heart disease (CHD) or combined cardiovascular disease (stroke-treated heart failure, CHD, coronary revascularization, angina, or peripheral arterial disease) during the study, despite the fact that the lisinopril group had baseline characteristics (ie, male sex and previous coronary revascularization) that predicted a higher risk of cardiovascular complications. One could conclude that R patients, who are inherently at higher risk for CHD, actually had similar outcomes to V patients with equal BP control, indicating that suppression of renin-angiotensin with lisinopril provided additional benefits.20 Although the interpretation of this post hoc analysis of ALLHAT patients who were controlled with monotherapy is limited, it suggests that alternative conclusions can be drawn from this trial that are more consistent with findings of previous studies and that contrast with the results originally reported by the ALLHAT investigators. Specifically, hypertension can often be controlled by monotherapy, although the first drug chosen is frequently not the best drug.14 Initial drug failure is often the consequence of incorrect targeting of the pathophysiologic mechanism (R-type v V-type) in the individual patient. Rather than being an indication for progression to multiple drug treatment, failure of the initial drug mandates that the ineffective agent be discontinued and an R-type drug substituted with a V-type drug or vice versa. Failure to consider these factors in the choice of antihypertensive drugs in the ALLHAT stepped care algorithm ultimately led to higher BP in the lisinopril treatment group. Accordingly, the unexpectedly higher risk of cardiovascular complications in the lisinopril group, described in the initial ALLHAT report, was abolished when monotherapy was correctly chosen.16,19 This suggests that, for the lisinopril group, the choice of redundant R-type drugs for both Step 1 and Step 2 confounded the initial analysis and placed these patients at a disadvantage. In summary, hypertensive patients can be stratified according to whether they have predominantly R-type or V-type mechanisms underlying their hypertension. This mechanistic approach increases the likelihood that long-term control of BP can be achieved with monotherapy. Failure to consider these factors in the choice of antihypertensive drugs can lead to higher BP and greater risk of cardiovascular complications. This pathophysiologic analysis of antihypertensive drug selection has important implications regarding optimal control of BP and reduction of cardiovascular risk.