Maximizing Diuretic Therapy in Resistant Hypertension

A 50‐year‐old man was referred to the hypertension clinic because of resistant hypertension of 2 years’ duration. He claimed adherence to a regimen that at the time of referral consisted of felodipine 10 mg twice daily, lisinopril 40 mg daily, furosemide 20 mg twice daily, terazosin 5 mg twice daily, potassium 10 mEq daily, and allopurinol 300 mg daily. Two years previously, sleep diagnostic testing performed because of daytime somnolence revealed moderate obstructive sleep apnea with 9 apneic and 7.2 hypopneic events/hour. Successful continuous positive airway pressure autotitration was completed and the patient noted improved daytime energy. He was sedentary and obese but had experienced a 10‐lb weight loss during the past month after cutting back on sodas. Nonetheless, swelling of his feet, which had occurred following an increase in the dose of felodipine, persisted. He denied any breathing difficulty or chest discomfort. There was a history of controlled, moderate persistent asthma and podagra.

Examination revealed a 177.8 cm in height individual with prominent central obesity, weighing 121.6 kg and having a blood pressure (BP) level of 172/62 mm Hg in both arms. He had 2+ ankle edema. Laboratory studies showed the following values: potassium 4.0 mmol/L (normal, 3.5–5.0 mmol/L), fasting glucose 98 mg/dL (normal, 70–99 mg/dL), creatinine 1.2 mg/dL (normal, 0.7–1.3 mg/dL), estimated glomerular filtration rate 50 cc/min, and uric acid 6.7 mg/dL (normal, 3.4–7.2 mg/dL).

Computed tomography of the abdomen performed 3 years previously for abdominal pain had showed normal adrenal glands and kidneys.

Felodipine dosage was reduced to 10 mg and furosemide was replaced with hydrochlorothiazide (HCTZ) 25 mg daily. Edema resolved and the follow‐up BP reading was 152/64 mm Hg. Sequential hydralazine was then added, going from 25 mg twice daily to 50 mg twice daily without any further BP reduction and with the return of bothersome edema. The 24‐hour urine metanephrine and catecholamine levels were normal; 24‐hour urine‐free Cortisol level was 10.7 μg/24 h (normal, 4.0–50 μg/24 h), a morning renin level was 37.7 ng/mL/h (normal, 0.65–5.0 ng/mL/h), and a morning aldosterone level was 7 ng/dL (normal, ≤28 ng/mL). Hydralazine was discontinued and the edema resolved.

The HCTZ was changed to chlorthalidone 25 mg; then spironolactone 12.5 mg was added and increased to 25 mg daily. A follow‐up BP reading was 126/68 mm Hg and potassium level was 3.7 mEq/L without potassium replacement. Figure 1 illustrates the BP response to treatment.

Figure 1.

. Blood pressure (BP) response to treatment. HCTZ indicates hydrochlorothiazide.

DISCUSSION

The most common cause of resistant hypertension is probably volume overload based on several factors including excess dietary sodium, ¹ chronic kidney disease with reduced ability to excrete a sodium load, and abnormal pressure natriuresis. ² Pressure natriuresis refers to the BP set point above which increased natriuresis occurs via renal mechanisms to maintain that BP. ³ Sodium retention may also occur as a secondary phenomenon resulting from certain antihypertensive medications, particularly sympatholytics and primary vasodilators, ⁴ , ⁵ as well as from medications other than those used to treat hypertension. This patient had an estimated glomerular filtration rate of 50 cc/min with stage 3 kidney disease secondary to underlying hypertensive nephropathy. Renal parenchymal disease is associated with increased plasma volume. ⁴ , ⁶ Obesity has also been associated with increased salt sensitivity, defined by an increased systolic pressure response to sodium loading. ⁷ The Table lists factors predisposing to volume overload.

It follows that medication manipulations to treat resistant hypertension involve maximizing and optimizing the diuretic regimen. The success of this strategy for treating resistant hypertension has been reported by 2 university hypertension clinics. At the Yale University School of Medicine (New Haven, CT) clinic, 50% of medical manipulations resulting in successful control involved adding a diuretic and 24% modified the diuretic therapy. ⁸ Modification of diuretic therapy included changing the class of diuretics from a loop diuretic to a thiazide when renal function was normal and vice versa when renal function was reduced. Fourteen years later, the experience of the Rush University Medical Center (Chicago, IL) hypertension clinic was similar: 58% of their cohort were believed to have drug‐related causes for their resistant hypertension; the most common cause was a suboptimal medication regimen. Sixty patients underwent a change in diuretic therapy: 21 patients, representing 35% of total medication changes achieving BP control, were prescribed a change in diuretic class; and 17 (28% of the total medication changes) had an increase in their diuretic dose ⁹ (Figure 2). Addressing drug‐related causes led to BP control in 64% of the patients, making these causes the most remediable forms of resistance. Other studies have also shown that increased plasma volume in patients with uncontrolled hypertension already receiving diuretic therapy may be successfully treated with additional diuretic therapy. ⁵ , ¹⁰

Figure 2.

. Changes in diuretic therapy in patients considered to have a suboptimal antihypertensive medication regimen. Reprinted with permission from Garg et al.

What is considered a maximal dose of a thiazide diuretic in patients with resistant hypertension? Approximately one‐half to two‐thirds of potential HCTZ responders will respond to 12.5 mg HCTZ and another 10% to 15% will respond to 25 mg HCTZ; 50 mg HCTZ will capture 80% to 90% of potential responders. ¹¹ With doses higher than 50 mg HCTZ, the dose response curve flattens such that minimal additional efficacy is associated with an increase in adverse metabolic side effects, primarily hypokalemia. Chlorthalidone, a thiazide‐like diuretic, is being used somewhat more frequently, particularly when a higher dose thiazide is required in patients with more resistant hypertension. Dose equivalency estimates, determined in only a few studies at the high end of the dose spectrum, ¹² are about 1:2, such that the BP reduction produced by 25 mg chlorthalidone approximates that of 50 mg HCTZ. Factors favoring chlorthalidone include (1) its usage as a thiazide class representative in successful hypertension treatment trials such as the Antihypertensive and Lipid‐Lowering Treatment to Prevent Heart Attack Trial (ALLHAT) and Systolic Hypertension in the Elderly Program (SHEP), and (2) its improved efficacy over 24 hours compared with HCTZ, ¹³ which would be predicted by its longer half‐life. Additionally, experience with resistant hypertension suggests that BP control may benefit from a switch from HCTZ to chlorthalidone. ¹⁴ Therefore, many practitioners prefer maximizing the thiazide dose at 25 mg chlorthalidone rather than 50 mg HCTZ for patients with resistant hypertension.

Prescribing spironolactone together with a thiazide diuretic is not a new concept, having been reported since 1962. ¹⁵ When used together for initial treatment of hypertension in this early work, as opposed to its use in resistant hypertension, the reported results were mixed. In the study by Wolf et al, ¹⁶ when spironolactone doses of 25 mg, 100 mg, and 200 mg daily were used with or without 100 mg HCTZ, the most effective BP lowering was in the group using 100 mg spironolactone alone. BP in this group decreased 20/9 mm Hg compared with 19/8 mm Hg when 100 mg HCTZ was added to spironolactone. Other studies showed additive benefit when HCTZ and spironolactone were used together compared with when used separately. ¹⁷ , ¹⁸ , ¹⁹ Hypokalemia occurs more commonly with chlorthalidone than with HCTZ; all of the combination studies with spironolactone showed significant improvement in hypokalemia. ¹⁵ , ¹⁶ , ¹⁷ , ¹⁸ , ¹⁹ , ²⁰ , ²¹ , ²² Adding spironolactone was more effective than using amiloride in correcting chlorthalidone‐related hypokalemia ²⁰ and more effective than using potassium chloride in raising serum potassium levels. ²¹ While spironolactone combined with an angiotensin receptor blocker (ARB) may cause serious hyperkalemia, ²³ such an interaction may be obviated when chlorthalidone is also given. In the present case, the combination of spironolactone, chlorthalidone, and an angiotensin‐converting enzyme (ACE) inhibitor resulted in normokalemia as well as hypertension control.

The present patient was an attractive spironolactone candidate. Even in the absence of evidence of primary aldosteronism, patients with resistant hypertension that is already treated with diuretics and ACE inhibitors or ARBs respond to spironolactone. ²⁴ In patients already taking antihypertensive drugs, the measurement of an aldosterone‐to‐renin ratio does not necessarily predict response to spironolactone. ²⁵ In this case, the high renin value was attributed to treatment with a diuretic and ACE inhibitor, and the aldosterone level was low. Another study showed that obesity and sleep apnea are frequently associated with inappropriate aldosterone levels and may be more responsive to spironolactone. ²⁶ This patient was obese and was adherent to nighttime continuous positive airway pressure (CPAP) because of quality of life improvement, but the CPAP did not reduce his BP.

What about estimating plasma volume of patients with resistant hypertension to determine candidacy for diuretic maximization? An individualized noninvasive assessment of hemodynamic status based on thoracic bioimpedance was found to outperform expert clinical judgment in one study. ¹⁰ The success of the bioimpedance‐directed protocol was attributed to more aggressive diuretic therapy in patients already receiving diuretics when testing showed that patients were volume expanded. More specifically, while the specialists with an empiric approach used thiazide diuretics more frequently, the bioimpedance group was prescribed loop diuretics more often by physicians willing to tolerate more medication‐related azotemia when directed by hemodynamic measurements. It would be of interest to learn in more detail the extent of the drug‐related azotemia that was tolerated and the baseline renal function of those patients most likely to benefit from the loop diuretic; it appears that spironolactone was not used in either patient group. In any case, thoracic bioimpedance is expensive, not generally available, and probably not necessary; increased diuretic therapy can be given empirically. Physical examination has also been suggested to assess volume status and direct intensification of diuretic therapy. ²⁷ Edema is a nonspecific sign, however, and when related to therapy with a calcium channel blocker, is generally not due to increased plasma volume and not responsive to diuretic therapy. In this case discussion, because of edema, the referring physician was misdirected from prescribing the more optimal thiazide therapy to a loop diuretic which failed to benefit either the edema or the hypertension. An initial hypertension clinic maneuver was to change the furosemide to HCTZ and reduce the felodipine dose which led to edema resolution and some BP improvement. However, it was only after a change to 25 mg chlorthalidone and later addition and up‐titration of spironolactone that good BP control was achieved.

Intensification of diuretic therapy with a goal of reducing plasma volume may be a misunderstanding. Significant reduction of plasma volume during the first few weeks of thiazide treatment is replaced by near normalization of plasma volume and reduction of peripheral vascular resistance, indicating that vasodilatation is the primary sustaining mechanism of BP improvement. In an early study examining plasma volume with I¹³¹‐labeled albumin in patients treated with spironolactone and chlorthalidone together, a significant decrease in plasma volume was not detected. ¹⁵ Prescribing spironolactone to address compensatory hyperaldosteronism caused by chlorthalidone may partially explain the additive BP‐Iowering effect when these 2 drugs are used together. The commonly used dose range of spironolactone in the absence of hyperaldosteronism is 12.5 to 50 mg. Another treatment option is combination pill therapy; doses of 25 mg or 50 mg spironolactone are available in combination with either 25 mg or 50 mg HCTZ. More than 60% of patients with resistant hypertension may have a response to increased diuretic therapy. ²⁸