Fixed‐Dosed Combinations Are Not Indicated as Initial Therapy: A Debate

Most patients with hypertension will require 2 or 3 antihypertensive agents to achieve blood pressure (BP) control, especially if there is evidence of target organ involvement.1 Patients with diabetes or chronic kidney disease may need 3 or 4 antihypertensives. Many authors promote the use of fixed-dose combinations (FDCs) for initial therapy for hypertension or for persons on multiple medications.2-4 The Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure (JNC 7)5 noted the importance of initiating 2 antihypertensives for stage 2 hypertension and acknowledged a role for FDCs. I believe that there is a rather limited role for FDCs. It is my opinion that initiating therapy with FDCs is usually unnecessary, may not be as effective as believed, is often not optimal therapy, and is generally more costly. I believe that FDCs should generally be reserved for use once a regimen is stable. Now that JNC 8 deliberations have begun, it will be important to better characterize the role of FDCs. Each of these issues will be addressed in this editorial. The Valsartan Antihypertensive Long-Term Use Evaluation (VALUE)6 reported that there was a significant reduction in risks of cardiovascular events when a systolic BP <140 mm Hg was achieved within 6 months, regardless of which drugs were used. The International Verapamil SR-Trandolapril (INVEST) study7 found that the more time a patient’s BP was <140/90 mm Hg, the lower the risk. Unfortunately, therapeutic or clinical inertia in which medications are not adjusted when BP is not controlled, is one of the most common causes of poor BP control.8, 9 Proponents of FDCs argue that BP control could be achieved more rapidly, and patient adherence and persistence could be improved with FDCs compared with multiple single medications. For example, the Avoiding Cardiovascular Events Through Combination Therapy in Patients Living With Systolic Hypertension (ACCOMPLISH) trial 10 compared the FDC of hydrochlorothiazide (HCTZ) and benazepril with the FDC containing amlodipine and benazepril. BP was controlled in 37% of patients at baseline. By 6 months, BP was controlled in 73% of all patients and in 43% of diabetics. Are the FDCs essential to achieve these high control rates? We recently completed a randomized controlled trial comparing the results of physician/pharmacist collaborative intervention with a control group.11 Patients in the control group were informed of their BP at each study visit. They received written BP education and their physicians were alerted to the lack of BP control. In this study, all patients had uncontrolled BP at baseline and many had been uncontrolled for long periods despite multiple medications (the average number of medications at baseline was 1.5). BP control was achieved in 89% of patients in the intervention group compared with 53% in the control group by 9 months (Figure 1). Most importantly in our study, BP was controlled in 82% of diabetic patients (<130/80 mm Hg) compared with 43% in ACCOMPLISH. Blood pressure (BP) control in the Collaborative Management of Hypertension Study. From Carter BL et al.11 While our study was small, we did not use forced titrations of specific drugs. Rather, the intervention simply involved tailoring single entity medications to patient needs to optimize therapy based on JNC 7. FDCs were rarely used in our study. Excluding combinations of diuretics with potassium-sparing agents, 2 (3%) patients in the control group were receiving FDCs at baseline, which increased to 6 (9%) at 9 months; these included diuretics plus a β-blocker (1), angiotensin-converting enzyme inhibitor (ACEI) (2), or angiotensin receptor blocker (ARB) (3). In the intervention group, 7 patients (7%) were taking an FDC at baseline; the number increased to 15 (15%) at 9 months (diuretics plus a β-blocker [1], ACEI [8] or ARB [5], or the combination of an ACEI/calcium channel blocker [CCB] [1]). To achieve high BP control rates, interventions made by pharmacists were performed quickly and follow-up was frequent. The vast majority of changes occurred in the first 2 months of the intervention and dose titrations usually occurred every 2 to 4 weeks.12 Thus, our study clearly demonstrated that high BP control rates, including in patients with diabetes, can be achieved without FDCs. Many authors argue that FDCs uniquely achieve high BP control rates because they simplify choices for physicians and increase convenience for patients. While there is evidence to support these concepts, it is impossible to compare forced drug titrations and scheduled research visits in a clinical trial such as VALUE or ACCOMPLISH with what will happen in real practice. It is possible that a physician who has been informed of the great success with FDCs will place a patient on an FDC and not make dosage adjustments, further contributing to clinical inertia. We actually see this problem quite frequently when patients are placed on an FDC such as a CCB or an ACEI. BP goal is sometimes not achieved for a very long period, when what was needed was the addition of a low-dose thiazide diuretic. Titration of FDCs can be cumbersome for busy primary care physicians. For instance, most of the ARB/HCTZ combinations have multiple doses of the ARB but the HCTZ dose is only available at 12.5 mg. This dose is a reasonable starting dose but it is too low for most patients, and outcome clinical trials show that the dose of HCTZ should be 25 to 50 mg daily.5 If BP is still not controlled with titration of the FDC, the physician must either add another drug, switch agents, or add a separate dose of HCTZ, all of which negate the benefit of the combination. This point also highlights the fact that nearly all of the FDCs that contain a diuretic have HCTZ, which is inferior as a 24-hour antihypertensive when compared with chlorthalidone.13, 14 While clinic BP may be similar with HCTZ 50 mg and chlorthalidone 25 mg, the important nighttime BP is lower with chlorthalidone.14 The 2 most effective strategies that were used in our Collaborative Management of Hypertension study11 for patients with previously uncontrolled hypertension was to switch to chlorthalidone or add low-dose spironolactone even in patients already taking a CCB plus an ACEI or ARB.12 The need for such add-on therapy greatly limits the presumed simplicity of the FDC. A recent study from the Mayo Clinic suggests that physicians and patients are reluctant to increase antihypertensive medications to more than 3 drugs.15 Their study also found that it was much more difficult to achieve BP control in older patients. Their innovative analysis still does not address the fact that these physicians were also very likely succumbing to clinical inertia that could be overcome by organizational change. Nonetheless, the clinician must always be concerned with pill burden and cost. So why then not use an FDC to limit pill burden? Nearly all health plans, including those involved with Medicare, have now moved to 3 tiers of cost sharing for pharmaceuticals. The largest health plan in our state places drugs such as verapamil/trandolapril, amlodipine/valsartan, or aliskiren/HCTZ in tier 3, which means the patient must pay 50% of the cost. These FDCs would cost the patient $40 to $54 per month depending on the agent. Irbesartan/HCTZ or valsartan/HCTZ are tier 2 drugs in this health plan and would cost the patient $22 to $28 per month. However, if the clinician prescribed 2 tier 1 drugs such as chlorthalidone and lisinopril, “the patient would pay nothing” since there is no co-pay for tier 1 drugs. In this health plan and many others, the physician could prescribe 4 drugs for a patient with resistant hypertension (chlorthalidone, lisinopril, amlodipine, and spironolactone as an example), and the patient “would also pay nothing.” The cost differences are even more dramatic if the patient pays for their own medications since they could get these 4 medications as generics. When considering the fact that many of these complicated patients are taking 1 or 2 lipid-lowering agents and perhaps several medications for other chronic conditions like diabetes or arthritis, cost becomes a major burden that may likely trump pill burden. The physician should arm themselves with cost data and have frank discussions with the patient to determine their desires to either address cost or pill burden. If the patient suggests that pill burden is a bigger concern, then perhaps in the above example the patient could be switched to HCTZ/spironolactone plus single agents of lisinopril and amlodipine (3 pills and still no cost to the patient) or to HCTZ/spironolactone plus amlodipine/benazepril generic (2 pills and no cost to the patient). In my experience, cost is a far bigger issue than pill burden. Most patients will not need the more expensive tier 3 FDCs, and there are enough inexpensive generic combinations to limit both pill burden and cost to the patient. The issue of possible altered pharmacokinetics or pharmacodynamics when 2 ingredients are combined in one dosage form is rarely discussed. After all, pharmaceutical companies study these parameters when compared with single-ingredient tablets administered at the same time. Why worry? Howes and colleagues 16 evaluated the pharmacokinetics of enalaprilat (the active metabolite of enalapril) and HCTZ in a 4-way comparison of a combination tablet of enalapril (10 mg)/HCTZ (25 mg) with a single dose of an enalapril tablet (10 mg), a single dose of an HCTZ tablet (25 mg), and simultaneous administration of separate tablets of enalapril (10 mg) and HCT (25 mg) in normotensive volunteers (n=12; 21–26 years of age). HCTZ was bioequivalent under all conditions. The bioavailability of enalapril was 12% lower when given with HCTZ as separate tablets and 19% lower when given as an FDC when compared with enalapril given alone. While these differences achieved borderline significance (P=.053 and P=.054, respectively), the study may have had low power with only 12 subjects. Even though these subjects were normotensive volunteers, BP decreased 20/15 mm Hg with enalapril alone, 19/16 mm Hg when given as separate tablets with HCTZ, but only 16/13 mm Hg when given as an FDC. Another bioavailability study in 12 healthy volunteers (aged 24–47 years) evaluated both atenolol and chlorthalidone following their administration as an FDC (“Tenoretic”), as 2 individual tablets, atenolol by itself and chlorthalidone alone, at doses of 100 mg atenolol and 25 mg chlorthalidone. There were no statistically significant differences between the 3 formulations of atenolol in terms of individual blood levels, half-life, area-under-the-curve values, and urinary excretion. The chlorthalidone blood levels were 10% higher and urinary excretion 24% higher following the administration of the FDC when compared with the free combination or the chlorthalidone tablet.17 These studies suggest that pharmacokinetics and probably pharmacodynamics may differ with FDCs when compared with individual tablets. These types of studies are always conducted in young, healthy normal volunteers, with very little information in the elderly. Pharmaceutical parameters such as tablet strength, tablet dissolution, excipients to bind the tablet, and the resulting absorption can change in an FDC compared with individual agents. But why does this matter if bioavailability and pharmacokinetics appear to be fairly similar when studied in healthy volunteers? Modest changes in bioavailability or even the rate at which a drug dissolves can have more dramatic effects on serum concentrations in the elderly due to altered gastrointestinal and hepatic enzyme activity with aging. For instance, we found that the bioavailability of 2 generic formulations of verapamil were nearly identical to the brand name drug in young patients.18 However, one generic product actually had greater serum concentration and area under the curve (77% and 43% higher, P<.02) compared with the branded verapamil in the elderly. This was associated with prolongations of the PR interval, suggesting a clinically relevant effect. These age-related differences were likely due to more rapid absorption of the generic products than the brand name drug when metabolic enzymes were overwhelmed in some elderly patients. We confirmed these findings in another study in which BP dropped more quickly with a different generic formulation in the elderly and serum concentrations were 11% higher with the generic in elderly patients.19 It is possible that formulating another drug with a CCB to create an FDC could change dissolution and/or absorption leading to very different pharmacokinetics in the elderly that might not be observed in the young. While the clinical relevance of such differences may be hard to determine, they certainly could make titrating to a BP response more challenging with an FDC. Figure 2 displays a proposed strategy for optimizing antihypertensive regimens.20 The clinician can enter the algorithm at various points depending on the existing regimen. For instance, if the patient’s BP is not controlled despite being on an ACEI and a CCB, chlorthalidone should be added. Chlorthalidone is listed as the preferred thiazide since it is more effective at controlling BP for the full 24-hour period without an increased risk of hypokalemia at the doses commonly used.13, 14 Strategy to optimize multidrug antihypertensive regimens.1 Adapted from Trewet and Ernst.20 A indicates angiotensin-converting enzyme inhibitor (ACEI) or angiotensin receptor blocker (ARB); B, β-blocker; C, calcium channel blocker; D, diuretic. 1All agents should be prescribed generically when possible to eliminate or minimize co-pays. 2ARBs should be reserved for intolerance to ACEIs. 3β-blockers should usually be reserved for patients with ischemic heart disease and/or heart failure and may be considered earlier in such patients. BP can be controlled in most patients treated with chlorthalidone plus a CCB plus an ACEI or ARB. It is also important to titrate medications every 1 to 4 weeks, depending on the patient, in order to achieve BP control within 6 to 9 months.6, 7 Many of these titrations can be conducted by telephone if the patient performs home BP monitoring. If BP is not controlled on such a regimen, the addition of spironolactone was extremely effective at achieving control in our study and others.21, 22 Endocrine adverse reactions are much less common when doses of spironolactone are 25 to 50 mg daily but eplerenone could be used if side effects cannot be tolerated with spironolactone.23 Less traditional regimens such as combining a dihydropyridine CCB with a nondihydropyridine CCB24, 25 may be considered if BP control still cannot be achieved. These combinations could be used with little or no cost for co-pays using our health plan as an example. These more complicated cases will likely require more time to achieve control but rapid titration is still important. Potential causes for resistant hypertension such as secondary hypertension, poor medication adherence, and/or dietary indiscretion should be considered.20, 26 Once the patient’s regimen has been stabilized, it is perfectly reasonable to switch some agents to FDCs. However, if the strengths available in the FDCs are different than those in the single products, BP will need to be monitored carefully after the switch. FDCs, especially expensive combinations of CCB/ACEI or ARB/HCTZ, are rarely needed to achieve BP control. Rational drug titrations with rapid BP control can be accomplished with single-entity products that cost the patient little or no money out of pocket if the patient has drug coverage. The issue of cost is even more dramatic if they do not have drug coverage or if they are in the Medicare Part D “doughnut hole.” If a regimen contains ≥3 antihypertensives, the physician should engage the patient in honest and frank discussions about strategies to limit the pill burden while still limiting the cost to the patient.