Hypertension is a common disease in our

AJH 2000;13:1161 1167 Effect of Indomethacin on Blood Pressure in Elderly People With Essential Hypertension Well Controlled on or Trefor O. Morgan, Adrianne Anderson, and Denise Bertram Arthritis and hypertension are frequent comorbidities in the elderly hypertensive population. Nonsteroidal anti-inflammatory drugs are often used to relieve pain in arthritic patients but a side effect is sodium retention and consequent elevation of blood pressure (BP). The effect of dihydropyridine calcium blocking drugs is relatively independent of sodium intake, whereas the angiotensin-converting enzyme (ACE) inhibitors effects can be blunted by a high-sodium diet. This study compared the effects of indomethacin with placebo in elderly patients with essential hypertension who had been controlled with amlodipine or enalapril. Indomethacin 50 mg twice daily or placebo was administered for 3 weeks in a double-blind crossover study to patients controlled with amlodipine or enalapril. The response was assessed by ambulatory BP measurement. Indomethacin raised BP and lowered pulse rates in patients taking enalapril but had little effect in patients receiving amlodipine. The difference caused by indomethacin between the two groups was 10.1/4.9 mm Hg increase in BP and a 5.6 beats/min fall in pulse in people taking enalapril. Addition of indomethacin to patients taking either drug caused a rise in weight and a fall in plasma renin. It is postulated that the effect is due to inhibition of prostaglandin synthesis, which causes sodium retention. In patients taking amlodipine, the fall in plasma renin ameliorates the effect of sodium retention on BP. In patients taking enalapril, plasma renin falls but this is not translated into an effect because of the blockage of converting enzyme. Thus, the full effect of sodium retention on BP is expressed. In patients treated with indomethacin, fewer patients may respond to ACE inhibitors. However, the major problem is the patient who intermittently takes indomethacin or other nonsteroidal anti-inflammatory drugs, which, if a person is treated by an ACE inhibitor causes BP to go out of control. In such patients amlodipine would appear to be a preferred choice to enalapril. Am J Hypertens 2000;13:1161 1167 2000 American Journal of Hypertension, Ltd. KEY WORDS:, enalapril, calcium blocking drugs, ACE inhibitors, indomethacin, nonsteroidal anti-inflammatory drugs, arthritis, hypertension, blood pressure. Hypertension is a common disease in our community and the prevalence increases as the population ages. In elderly people arthritis and other inflammatory problems become more frequent and a significant number of people are treated with nonsteroidal anti-inflammatory drugs (NSAID). NSAID cause salt and water retention and this may be the reason they elevate blood pressure (BP) in both normotensive and hypertensive people. 1 3 If a person is taking an NSAID and is being Received November 12, 1999. Accepted April 4, 2000. From the Department of Physiology, University of Melbourne and Hypertension Clinic, ARMC, Heidelberg, Australia. This study was supported by Pfizer Pty. Limited. Address correspondence and reprint requests to Prof. T. Morgan, Department of Physiology, University of Melbourne, Parkville, Victoria 3052, Australia; e-mail: t.morgan@physiology.unimelb.edu.au 2000 by the American Journal of Hypertension, Ltd. 0895-7061/00/$20.00 Published by Elsevier Science, Inc. PII S0895-7061(00)01204-8

1162 MORGAN ET AL AJH NOVEMBER 2000 VOL. 13, NO. 11 treated for hypertension the dose of the drug will be increased or the drug altered so that the hypertension is controlled. However, if a person intermittently takes a NSAID the BP may be controlled when they are not taking the NSAID and go out of control when NSAID is used. People taking a NSAID may not respond to certain antihypertensive drug classes as well as to others. When therapy is started with most classes of antihypertensive drugs, some sodium retention may develop because of BP reduction. This may lead to a person becoming relatively insensitive to the drug and the addition of a diuretic frequently allows control to be maintained. The response to the angiotensin-converting enzyme (ACE) inhibitors is relatively sensitive to the sodium intake of an individual. 4 The lower the sodium intake, the greater the fall in BP when an ACE inhibitor is used. This probably relates to the compensatory rise in renin and hence angiotensin II that occurs when sodium restriction is introduced. 5,6 This in turn blunts the response to sodium restriction and then blockade of this system causes a fall in BP. With the calcium-channel blocking drugs the response appears to be independent of sodium chloride intake and in this regard they differ from most other antihypertensive drugs. 7 In addition, diuretics appear to have little additional effect when a person is taking a calcium-channel blocking drug. 8,9 In view of the fairly extensive association of arthritis and hypertension, this study was undertaken to see if people responded in a different way when a NSAID was used according to the class of drug they were taking. The two classes of drugs that were studied were an ACE inhibitor (enalapril) and a dihydropyridine calcium-channel blocking drug (amlodipine). The NSAID used was indomethacin, as this compound has clearly shown elevations in BP with its administration. 3,10 METHODS The patients recruited to the study were men and women, age 45 to 85 years, with essential hypertension well controlled taking either 20 or 40 mg/day enalapril or 5 or 10 mg/day amlodipine. This control was assessed by BP measured 24 h after drug administration. The patients entered a preliminary 2-week run-in period during which they received twice-daily placebo matching indomethacin. They continued taking their normal dose of enalapril or amlodipine. The enalapril or amlodipine was taken between 6:00 am and 9:00 am. On the day of their visit to the clinic they did not take their antihypertensive medication. Supine BP was recorded at the time of entry and at Week 2 of the run-in period. The BP was required to be lower than 160/90 mm Hg and the supine diastolic BP should not have differed by 5 mm Hg between the two consecutive visits. People who met these criteria were then randomized to the study. The study consisted of two parallel groups, one receiving enalapril and the other receiving amlodipine. Patients who met the entry criteria were randomized to a double-blind crossover study in which they received placebo, 1 capsule twice daily, or indomethacin, 50 mg twice daily. Each study period was 3 weeks in duration. Each patient was seen at Week 2 and Week 3 of each study period. At Week 3 of the study the patient had an ambulatory BP monitor fitted, was given study medication, and was seen the following day. The ambulatory BP monitoring continued for 26 to 27 h in most patients. The following patients were excluded from the study: patients with evidence of a cardiovascular or cerebrovascular episode within the previous 6 months or with unstable angina; patients in cardiac failure or heart block; patients with a history of peptic ulceration or gastrointestinal bleeding; patients with renal or hepatic dysfunction. Patients were not allowed to be receiving other antihypertensive drugs or NSAID. However, a dose of aspirin up to 650 mg daily was permitted. Informed consent was received from each patient and the study was approved by the Ethics Committee of the Austin & Repatriation Medical Centre. At each visit the patients were weighed, the BP was recorded in triplicate after 10 min supine and 5 min erect; a mean value was recorded for each. Blood was taken at the end of the run-in and at the end of each treatment period. Routine biochemistry was performed on this blood, but in addition, plasma renin and aldosterone were measured. The ambulatory BP was performed over a 26- to 27-h period using A&D model TM 2421. Blood pressure was taken every 15 min day and night and standard exclusion criteria were applied to the BP. The ambulatory BP was divided into the 24-h mean, the daytime mean, the nighttime mean, and the morning mean. The daytime mean was taken as the interval between 6:00 am and 10:00 pm, the nighttime mean as the BP between 10:00 pm and 6:00 am, and the morning mean as the 2-h period starting after the patient got out of bed. Power calculations indicated that to detect a 5-mm Hg difference in supine systolic BP between patients taking indomethacin or placebo, assuming a standard deviation of 8 mm Hg and a power of 80% required a total of 44 patients, 22 taking amlodipine and 22 taking enalapril. The prime response analyzed was the change in 24-h ambulatory BP after the addition of indomethacin to treatment with amlodipine compared to the change after its addition to treatment with enalapril. In addition to this primary analysis, changes in the daytime mean, the nighttime mean, and the morning mean of

AJH NOVEMBER 2000 VOL. 13, NO. 11 INDOMETHACIN AND HYPERTENSION 1163 TABLE 1. FLOW PATTERN* Entered run-in 33 32 Randomized 30 31 Withdrawn Adverse events 3 4 Patient s request 3 1 Completed placebo and indomethacin 24 26 Two satisfactory ABPM 18 24 Protocol violation 0 1 Clinic analysis 24 25 ABPM analysis 18 23 ABPM ambulatory blood pressure monitoring. * Primary analysis, change in 24-h mean blood pressure caused by indomethacin in patients taking enalapril or amlodipine. TABLE 2. DEMOGRAPHICS OF PATIENTS WHO COMPLETED THE PROTOCOL AND HAD SATISFACTORY ABPM TAKING PLACEBO AND INDOMETHACIN Group Group Age; mean range (yr) 72; 55 88 69; 44 80 Gender (M/F) 15/3 18/5 Weight (kg) 76.9 77.8 Height (cm) 169.3 171.1 Years of hypertension 17.2 13.6 Systolic blood pressure (mm Hg) 147.6 152.7 Diastolic blood pressure (mm Hg) 77.2 81.0 Pulse (beats/min) 73.1 69.4 Abbreviation as in Table 1. the ambulatory BP were analyzed. Secondarily, the effect inside the two groups of placebo compared to indomethacin was assessed by the null hypothesis. RESULTS Sixty-five patients entered the run-in phase of the study; 30 of 33 patients taking enalapril and 31 of the 32 patients taking amlodipine were randomized to placebo or indomethacin. Three patients in both the enalapril and amlodipine group withdrew because of side effects and three people taking enalapril and one taking amlodipine were withdrawn at their request. Twenty-four of the patients taking enalapril completed both phases of the study but only 18 of these had a satisfactory ambulatory BP monitor on two occasions. Twenty-six patients receiving amlodipine completed the study and 24 of these had satisfactory ambulatory BP monitors on both occasions. The main reason for unsatisfactory ambulatory BP recordings was failure to meet standard preset requirements. One patient in the amlodipine group had a protocol violation by receiving prednisolone during the study and was accordingly excluded from the analysis (Table 1). The analysis based on clinic values was accordingly done on 24 patients taking enalapril and 25 taking amlodipine. The ambulatory BP analysis (the primary analysis) was done in 18 patients taking enalapril and 23 taking amlodipine. In a double-blind crossover study without values determined at the end of a run-in phase it is only possible to do the analysis on people who complete both arms. The demographics of these patients are shown in Table 2 and it can be seen that the patient groups were similar. Two patients were slightly outside the age entry criteria but were included for analysis purposes. The clinic BP of the patients who completed the full protocol is shown in Table 3. In the patients receiving enalapril, indomethacin caused a significant rise in supine systolic BP and a trend for a rise in the erect systolic BP. In the patients taking amlodipine, there was no significant difference in the BP between those taking indomethacin or placebo. There were also no significant differences in the supine pulse rate but there was a fall in the erect pulse rate in the patients taking enalapril when they received indomethacin. The ambulatory BP data are presented in Table 4. In the patients receiving enalapril, indomethacin caused a significant rise in the 24-h mean systolic and diastolic BP, the daytime and nighttime BP (Figs. 1, 2). There was a trend for the morning BP to rise, which was not significant for systolic, but was for diastolic BP (Table 4). There was a significant fall in pulse rate in the patients receiving indomethacin while taking enalapril. In the patients taking amlodipine, when they were given indomethacin there were no significant changes in the systolic or diastolic BP or pulse rate during any of the periods analyzed. The change in 24-h mean systolic BP in patients taking amlodipine receiving indomethacin was less than that in those patients taking enalapril receiving indomethacin (Table 4, Fig. 1). In addition to the primary analysis being positive the other analyses reached significance or had a trend for all BP measured. Fall in pulse rate in the patients receiving enalapril was greater than in those taking amlodipine. The mean changes in 24-h BP caused by indomethacin in patients taking enalapril and amlodipine are shown in Fig. 1 and are highly significant. The response of the individuals is indicated in Fig. 3 and the responses to indomethacin in people taking amlodipine are distributed around 0. Most of the patients taking enalapril had a rise in BP with indomethacin and in some patients this rise was quite marked. The weight in both groups of patients increased by

1164 MORGAN ET AL AJH NOVEMBER 2000 VOL. 13, NO. 11 TABLE 3. EFFECT OF INDOMETHACIN ON CLINIC BLOOD PRESSURE IN PATIENTS TAKING ENALAPRIL AND AMLODIPINE (n 18) (n 23) Placebo Indomethacin P Placebo Indomethacin P P, v Supine Systolic 150 10 160 10.02 153 15 157 14.19 0.31 Diastolic 79.8 7 80.8 7.62 80.8 7 82.0 9.59 1.00 Pulse 72.7 13 71.4 13.30 71.5 10 70.4 11.76 0.40 Erect Systolic 154 11 159 9.06 153 18 159 17.07 0.85 Diastolic 84.2 12 85.2 11.42 85.6 9 85.5 10.76.38 Pulse 79.7 14 75.8 13.02 78.4 11 76.7 11.87 0.05 Results are given as mean S.D. These data are for the patients who satisfactorily completed the ambulatory blood pressure monitoring on both occasions. The P value in the far left column refers to the comparison of the change in blood pressure in people taking enalapril with the change in those taking amlodipine. approximately 1 kg when they were given indomethacin. The plasma renin fell in both groups, probably indicating sodium retention. Patients taking enalapril had a higher plasma renin than those patients taking amlodipine while taking placebo. There were changes in the aldosterone level that were of marginal significance (Table 5). Side Effects The occurrence of drug-related side effects was similar in each group, with six patients complaining of side effects (Table 6). The most common effects were gastrointestinal, with abdominal pain, indigestion, and nausea being the most common. This caused the drug to be stopped in a number of the patients early in the study and these patients data were not analyzed. The side effects that caused cessation of therapy are indicated in Table 6. DISCUSSION The patients in both arms of the study (enalapril and amlodipine arm) were relatively similar in their demographics and characteristics though it is feasible that the drug had been selected for specific reasons in these patients and thus they were not randomly matched. Indomethacin caused a rise in the ambulatory BP measurements in patients taking enalapril and associated with this rise was a fall in pulse rate, possibly as a consequence of the rise in BP activating the baroreflex. Conversely, in the patients controlled tak- TABLE 4. BLOOD PRESSURE WHILE TAKING PLACEBO AND INDOMETHACIN RECORDED BY AMBULATORY MONITORING IN PATIENTS TAKING ENALAPRIL OR AMLODIPINE (n 18) (n 23) Placebo Indomethacin P Placebo Indomethacin P P, v 24-h mean Systolic 134 146.002 141 142.53.017 Diastolic 73 78.002 77 77.66.017 Pulse 75 69.0005 74 74.55.08 Daytime mean Systolic 137 149.003 143 145.27.03 Diastolic 75 80.007 78 79.30.07 Pulse 77 70.0002 76 74.34.05 Nighttime mean Systolic 118 130.002 128 130.53.02 Diastolic 62 69.0004 68 68.48.01 Pulse 67 63.07 67 66.75.22 Morning mean Systolic 145 155.10 149 149.61.10 Diastolic 79 86.04 82 81.57.06 Pulse 82 78.09 79 81.34.07

AJH NOVEMBER 2000 VOL. 13, NO. 11 INDOMETHACIN AND HYPERTENSION 1165 FIG. 1. The 24-h systolic blood pressure profile in patients on enalapril or amlodipine when given placebo and indomethacin. BP blood pressure. FIG. 2. The effect of indomethacin on 24-h ambulatory blood pressure in hypertensive patients controlled on enalapril or amlodipine. Abbreviation as in Fig. 1. FIG. 3. Distribution of alteration in 24-h mean ambulatory systolic and diastolic blood pressure in patients on enalapril or amlodipine who were given indomethacin. In this figure, all data have been included even if ambulatory blood pressure monitoring over 24 h was not satisfactory. In such cases, the change in blood pressure over the comparable recordings on the two traces have been compared (n 23 enalapril, n 26 amlodipine). Abbreviation as in Fig. 1. ing amlodipine there was no rise in BP and no alteration in pulse rates. The differences between the two groups were significant. This difference was significant not only for the mean 24-h BP but also for the daytime and nighttime means. It was not significant for the morning rise in BP; this possibly relates to the greater variation that occurs in this BP because of the inconsistency of the time of arising and thus inclusion of a wider range of values. Results with clinic BP were not as conclusive and this probably relates to the fewer number of measurements. However, both at Week 2 (not shown) and Week 3 of the study there was a significant rise in BP in patients taking enalapril treated with indomethacin, whereas the change in BP in those patients taking amlodipine treated with indomethacin was not significant. However, the difference between the two groups did not reach significance. In a small previous study, we demonstrated that indomethacin elevated clinic BP in people controlled by taking enalapril but not in those controlled taking felodipine. 11 This study supports that conclusion but extends the observation to demonstrate that there is a significant effect on the various ambulatory BP measurements and demonstrates that well-controlled individuals may go out of control. Similar conclusions were reached comparing the effect of indomethacin on people controlled taking nifedipine GITS and enalapril. 10 A recent study inferred that this effect was not seen when indomethacin was given to people controlled taking an AT 1 receptor blocking drug com-

1166 MORGAN ET AL AJH NOVEMBER 2000 VOL. 13, NO. 11 TABLE 5. EFFECT OF INDOMETHACIN ON WEIGHT, RENIN, AND ALDOSTERONE LEVELS Placebo Indomethacin P Placebo Indomethacin P Weight (kg) 78.2 12.1 79.1 12.5.058 77.7 11.5 78.6 12.2.03 Renin (ng/ml/h) 6.65 6.85 2.95 2.35.02 1.91 1.62 1.62 1.12.05 Aldosterone (pmol/l) 232 195 162 112.048 250 230 248 200.7 Mean SD. pared to placebo. 12 However, there were only eight people in the study and the power to show an effect was small. The mechanism of the rise in BP with indomethacin in both untreated and treated hypertensive patients is unclear. Indomethacin is a nonselective inhibitor of cyclooxygenase. It interferes with the synthesis of prostaglandins and this has an effect on the kidney, causing sodium retention. 2,13,14 It may also possibly have an effect on blood vessels, causing lack of vasodilatation and hence a rise in BP. Our hypothesis is that the most likely cause is explained by the weight gain, which indicates sodium retention. 11 In patients receiving an ACE inhibitor plasma renin is initially high because of inhibition of the short feedback loop caused by angiotensin II. However, in people taking an ACE inhibitor this high plasma renin has minimal effect on BP as there is a block to formation of angiotensin II. In both people taking enalapril and amlodipine there was a fall in plasma renin caused by sodium retention. In a normal person or an untreated person this fall in plasma renin activity and hence angiotensin II level ameliorates the rise in BP caused by sodium retention and volume expansion, in the same way as the response to sodium restriction and diuretics is blunted by the rise in plasma renin. 6 In the patients taking enalapril, even though there is a fall in renin TABLE 6. TREATMENT-RELATED ADVERSE EVENTS AND WITHDRAWALS Placebo Indomethacin Placebo Indomethacin Indigestion 0 1 (1) 0 3 (2)* Abdominal pain 0 2 0 3 (2)* Vomiting 0 1 (1) 0 0 Tachycardia 1 (1) 0 0 0 Cough 0 0 0 1 Nausea 0 1 (1) 0 0 Headache 0 0 0 1 No. of patients 1 (1) 5 (3) 0 (0) 6 (3) Number withdrawn in (). *Some patients had one or more symptoms. levels there will be no significant alteration in angiotensin II and thus the rise in BP from sodium retention is unchecked. In patients receiving a calcium-channel blocking drug the fall in plasma renin and the subsequent fall in angiotensin II means that the full effect of sodium retention is not seen as it is compensated for by the fall in renin and thus angiotensin II. In addition, people taking calcium-channel blocking drugs have a relative insensitivity to sodium intake and this may be another explanation. 7,8 Polónia et al 10 had similar results related to weight gain but because the increase in weight was similar in both groups they argued that the mechanism would not be caused by sodium retention but by some other mechanism. They ignored the crucial role that alterations in renin might play though the data in their study related to renin are confusing as the plasma renin in patients taking enalapril is only slightly elevated. Patients who respond to an ACE inhibitor usually also appear to response to a calcium-channel blocking drug but the reverse is not true. 15 This is probably because angiotensin II acts only on the receptor-mediated Ca 2 channel and unless there is angiotensin II present it is unlikely that ACE inhibitors will have a major effect. Calcium-channel blocking drugs, on the other hand, act on the voltage-regulated channel and this will allow a response to occur in many more patients than with the ACE inhibitor drug. From this study we would conclude two major points. First, if a person is taking a NSAID it is likely that they will respond better to a dihydropyridine calcium-channel blocking drug than to an ACE inhibitor. This point was not evaluated explicitly in this study but is a logical conclusion. However, the main point is that if a hypertensive person receives an NSAID intermittently it would be preferable to control them with a calcium-channel blocking drug rather than an ACE inhibitor, as the control is not abrogated during their NSAID medication. ACKNOWLEDGMENTS We acknowledge the cooperation of the patients and the nursing staff at the Hypertension Clinic at the ARMC Heidelberg.

AJH NOVEMBER 2000 VOL. 13, NO. 11 INDOMETHACIN AND HYPERTENSION 1167 REFERENCES 1. Brown J, Dollery C, Vades G: Interaction of non-steroidal anti-inflammatory drugs with antihypertensive and diuretic agents: control of vascular reactivity by endogenous prostanoids. Am J Med 1986; 81(suppl 2B):43 57. 2. Watkins J, Abbott EC, Hensby CN, Webster J, Dollery CT: Attenuation of the hypotensive effect of propranolol and thiazide diuretics of indomethacin. Br Med J 1980;281:702 705. 3. Chalmers JP, West MJ, Wing LMH, Bune AJC, Graham JR: Effects of indomethacin, sulindac, naproxen, aspirin, and paracetamol in treated hypertensive patients. Clin Exp Hypertens [A] 1984;6:1077 1093. 4. Morgan TO, Anderson A, Wilson D, Murphy J, Nowson C: The effect of perindopril on blood pressure in humans on different sodium intakes. J Cardiovasc Pharmacol 1987;10(suppl 7):S116 S118. 5. Morgan TO, Creed R, Hopper J: Factors that determine the response of people with mild hypertension to a reduced sodium intake. Clin Exp Hypertens 1986;A8: 941 962. 6. Anderson A, Morgan TO, Wilson D, Nowson C, Murphy J, Snowden R: Use of sodium restriction and enalapril in persons with moderate to severe hypertension. Med J Aust 1987;146:149 153. 7. Morgan TO, Anderson A, Wilson D, Myers J, Murphy J, Nowson C: Paradoxical effect of sodium restriction on blood pressure in people on slow channel calcium blocking drugs. Lancet 1986;1:793. 8. Morgan TO, Anderson A: Interaction of slow channel calcium blocking drugs with sodium restriction, diuretics and converting enzyme inhibitors. J Hypertens 1988; 6(suppl 4):S652 S654. 9. Cappuccio FP, Markandu ND, Tucker F, Sagnella GA, MacGregor GA: Does a diuretic cause a further fall in blood pressure in hypertensive patients already on nifedipine? J Hypertens 1987;5:733 738. 10. Polónia J, Boaventura I, Gama G, Camões I, Bernardo F, Andrade P, Nunes JP, Brandao F, Cerqueira-Gomes M: Influence of non-steroidal anti-inflammatory drugs on renal function and 24h ambulatory blood pressurereducing effects of enalapril and nifedipine gastrointestinal therapeutic system in hypertensive patients. J Hypertens 1995;13:925 931. 11. Anderson A, Morgan TO: Interaction of indomethacin with felodipine and enalapril. J Hypertens 1993; 11(suppl 5):S338 S339. 12. Olsen ME, Thomsen T, Hassager C, Ibsen H, Dige- Petersen H: Hemodynamic and renal effects of indomethacin in losartan-treated hypertensive individuals. Am J Hypertens 1999;12:209 216. 13. Laiwah ACY, Mactier RA: Antagonistic effect of nonsteroidal anti-inflammatory drugs on furosemide-induced diuresis in cardiac failures. Br Med J 1981;283:714. 14. Ferri C, Bellini C, Piccoli A, Carlomagno A, Bonavita MS, Santucci A, Balsano F: Enhanced blood pressure response to cyclooxygenase inhibition in salt sensitive human essential hypertension. Hypertension 1993;21: 875 881. 15. Morgan TO, Anderson A: Haemodynamic comparisons of enalapril and felodipine and their combination. Kidney Int 1992;41(suppl 36);S78 S81.