Nonsteroidal Anti-Inflammatory Drugs and Antihypertensives MARK C. HOUSTON, M.D., NasM/e, Tennessee

Approximately 60 million people in the United States have hypertension (BP 2140/90 mm Hg), 40 million have arthritis clinically suitable for nonsteroidal anti-inflammatory drug (NSAID) therapy, and millions take NSAIDs for non- arthritic conditions, creating considerable po- tential for concomitant administration of NSAIDs and antihypertensive agents. It is esti- mated that more than 20 million people are on concurrent therapy.

High-risk patients treated with NSAIDs should be identified and have blood pressure, renal function, and serum potassium fre- quently monitored.

Most NSAIDs produce mild elevations of nor- mal blood pressure levels and can partially or completely antagonize the effects of many an- tihypertensive drugs. The effect on blood pres- sure can vary from no effect to hypertensive crisis. In pooled studies, the average increase in mean arterial pressure was 10 mm Hg, and duration was short-lived or chronic. Significant interactions occur in about 1% of patients per year. The risk is greatest in the elderly, blacks, and patients with low-renin hyperten- sion. NSAIDs may block the antihypertensive effects of thiazide and loop diuretics, j?- adrenergic blockers, cu-adrenergic blockers, and angiotensin-converting enzyme inhibitors. No interactions have been reported with centrally acting (Y agonists or the calcium channel blockers.

T he potential for concurrent administration of drugs used in treating arthritis and antihyper-

tensive agents is considerable. Approximately 30% of the 40 million patients who have arthritis clini- cally suitable for nonsteroidal anti-inflammatory drug (NSAID) therapy are also being treated for hypertension. Thus, there are approximately 12 million patients in the United States receiving both NSAID and antihypertensive therapy. With the proliferation of NSAIDs and their use in non- arthritic conditions, the frequency with which the two types of agents are used concurrently is even higher. The true figure of concomitant usage proba- bly approximates 20 million patients.

The mechanism of the hypertensive effects of NSAIDs seems primarily related to their ability to block the cycle-oxygenase pathway of arach- idonic acid metabolism, with a resultant de- crease in prostaglandin formation. The prosta- glandins are important in normal modulation of renal and systemic vascular dilatation, glo- merular filtration, tubular secretion of salt and water, adrenergic neurotransmission, and the renin-angiotensin-aldosterone system. Blockade of salutary effects of prostaglandins by NSAIDs results in a complex series of events culminating in attenuation of the ef- fects of many antihypertensive agents.

In the last 15 years there has been growing awareness and concern that most NSAIDs are ca- pable of producing elevations in blood pressure in both normal individuals and in patients with treated and untreated hypertension and thus are capable of partially or completely antagonizing the effects of some antihypertensive agents. The mag- nitude of the blood pressure elevations varies among patients, ranging from little or no effect to hypertensive crises [l]. The average increase in mean arterial pressure has been reported as 10 mm Hg, and the effect may be short-lived or chronic [2]. This relatively minor increase probably would be of little clinical importance over a short period, but it may have considerable clinical significance if it per- sists over a prolonged time. In a study [3] of more than 2,000 patients receiving NSAIDs and antihy- pertensive agents, the frequency of drug interac- tions causing significant blood pressure elevations was reported as 1%.

NSAiDs AND ANTIHYPERTENSIVE AGENTS

Supported by a grant from G.D. Searle. Requests for reprints should be addressed to Mark C. Houston, M.D., Clinical

Associate, Professor of Medicine, St. Thomas Medical Plaza East, 4230 Harding

Indomethacin is the NSAID that has been most extensively evaluated for its potential to antagonize antihypertensive therapy. It has been shown to at- tenuate the antihypertensive effects of the thia- zides [4-61, loop diuretics [7], /3-adrenergic blockers [4,5,8-111, a-adrenergic blockers [121, and the an-

5A-42S May 17, 1991 The American Journal of Medicine Volume 90 (suppl 5A)

SYMPOSIUM ON CALCIUM ANTAGONISTS I HOUSTON

Figure 1. The nonsteroidal anti-inflammatory drugs.

Choline Salicylate (Antropan)

I Salicylates

I

Propionic Acids

Mefenamic Acid (Pqnstel)

I 2, g, z ‘Currently not available in the U.S.

giotensin-converting enzyme (ACE) inhibitors [13- 161. A relatively limited number of studies have accompanied the proliferation of the newer NSAIDs, and the indications are that most, if not all, share some of the attenuating action of indo- methacin on the antihypertensive agents. Figure 1 lists the NSAIDs currently available, and Table I lists some NSAIDs and the antihypertensive agents with which they are known to interact.

Several mechanisms whereby NSAIDs may con- tribute to hypertension and attenuate antihyper- tensive drug activity have been described. The chief mechanism, however, appears to be related to their effect of inhibiting the cyclo-oxygenase path- way of arachidonic metabolism (Figure 2). This re- sults in a decrease in the biosynthesis of all pros- tanoids, including the E and F series as well as thromboxane AZ. Thus, the actions of antihyperten- sive agents that at least partially exert their effects via renal prostaglandin mechanisms (i.e., the thia- zides and loop diuretics, P-adrenergic blockers, a- adrenergic blockers, and ACE inhibitors) would be attenuated; agents whose antihypertensive action is secondary to other mechanisms (e.g., calcium channel blockers and centrally acting drugs) would not be expected to have their effects antagonized by NSAIDs.

EFFECTS OF INDOMETHACIN Durao et al [81 were among the first to study the

effects of indomethacin-induced inhibition of endog-

enous prostaglandin synthesis on antihypertensive drug activity. They studied the effects of indometh- acin on seven hypertensive patients receiving the P-blocking agents pindolol(15 mgid) or propranolol (80-160 mgid). In four test phases of 10 days each, patients received /3 blockers alone, placebo, j3 blocker plus indomethacin (100 mgid), and p block- ers alone. Mean diastolic blood pressure rose signif- icantly from the first p blocker phase to the placebo phase, remained elevated in the p blocker-plus- indomethacin phase, and fell to initial levels in the second p blocker phase. Diastolic blood pressure values obtained in the p blocker-plus-indomethacin phase were not significantly higher than pressures

2 NSAlDs that Interact with Antihypertensive Drugs

lndomethacin

Flurbiprofen

Diclofenac Aspirin Naproxen Ibuprofen

Thiazides, furosemide, p blockers, IY blockers, ACE inhibitors

p blockers, piroxicam, thiazides, furosemde, p blockers (propranolol)

Hydralazine p blockers, spironolactone Thiazides,P blockers (tlmolol) Thiazides, p blockers

All NSAlDs and ASA reduce effects of: Thiazides Loop diuretics Spironolactone

-he data on sulindac is controversial (see text). ACt = angiotensin-converting enzyme; ASA = acetylsalicylic acid (aspirin); NSAlDs = nonsteroldal anti-inflammatory drugs.

May 17, 1991 The American Journal of Medicine Volume 90 (suppl 5A) 5A43S

SYMPOSIUM ON CALCIUM ANTAGONISTS / HOUSTON

Membrane phospholipids

I Phospholipase

4

/ :::I:::: borne p45o Lipoxygenase

1

I

+ Epoxygenase

HETE6

Leukotrienes

EETs

PGEp Thromboxane AC PGE2

yrglay 1

ptG” I Figure 2. The metabolism of arachidonic acid.

PGF, TxB2 DHETs DHET = dihydroeicosatrienoic acid; EET = epoxyeicosatrienoic acid; HETE = hydroxyeico-

HETE = hydroxyeicosatetraenoic acid; EET I epoxyeicosatrienoic acid; DHET E dihydroeicosatrienoic acid satetraenoic acid; PG = prostaglandin; Txi3, =

_ thromboxane B2.

obtained in the placebo phase, but were signifi- cantly different from values measured in both /? blocker phases (p <O.Ol).

Watkins et al [5] studied the manner in which in- domethacin attenuates the hypotensive effects of propranolol and thiazide diuretics. Indomethacin (100 mg/d) or placebo was administered for 3 weeks in a double-blind fashion to 15 hypertensive pa- tients, eight of whom were receiving propranolol and seven, thiazide diuretics. Blood pressure and the urinary excretion of prostaglandin Fs, an indi- cator of renal prostaglandin activity, were moni- tored. Compared with placebo, adding indometha- tin to the patients’ established antihypertensive regimen increased blood pressure by 14-15 mm Hg supine and 16-19 mm Hg erect in patients receiving propranolol, and by 13-19 mm Hg supine and 16-19 mm Hg erect in patients receiving thiazide diuret- ics (p ~0.05). The excretion of the major urinary metabolite of prostaglandin Fz was reduced by 67% in the propranolol-treated patients and by 57% in patients receiving a thiazide diuretic. There were no significant changes in body weight, creatinine clearance, urinary sodium excretion, or packed cell volume in either treatment group.

EFFECTS OF OTHER NSAlDs Wong et al [3] compared the effects of placebo

and three NSAIDs-sulindac, piroxicam, and na- proxen-on blood pressure and blood and urine lev- els of cyclooxygenase pathway metabolites in 20 hypertensive patients. Each patient’s blood pres- sure was stabilized on a twice-daily regimen of the p blocker timolol and a combination diuretic con- taining 50 mg hydrochlorothiazide and 5 mg amilo- ride. In a double-blind fashion, each patient was randomized to receive 4 weeks of each of the follow- ing: placebo, 1 tablet twice daily; piroxicam, 10 mg

twice daily; naproxen, 250 mg twice daily; and su- lindac, 250 mg twice daily. At the end of each 4-week therapeutic regimen, urinary concentra- tions of 6-keto-prostaglandin F1 (6-keto-PGF,), the stable hydrolysis product of prostacyclin (PGIa) and thromboxane Bz (TxBa) were measured, as were creatinine clearance, 24-hour urinary sodium excre- tion, glomerular filtration rate, and plasma renin activity. Plasma levels of 6-keto-PGFi and serum thromboxane levels were also measured.

There were no differences between the four groups in weight, urinary sodium, serum creati- nine, creatinine clearance, glomerular filtration rate, or plasma renin activity. All three NSAIDs caused significant and equivalent reduction in platelet thromboxane synthesis and of plasma 6-keto-PGF, concentration. However, sulindac did not differ from placebo with respect to renal syn- thesis of prostaglandins, whereas naproxen and pir- oxicam significantly reduced urinary TxBa and 6-keto-PGF,. Both supine and standing diastolic blood pressures were significantly lower with sulin- dac than with placebo and were significantly higher with piroxicam and naproxen than with sulindac.

The studies by Watkins et al [5] and Wong et al [3] illustrated the relationship between NSAID blockade of the cycle-oxygenase pathway and inhi- bition of renal synthesis of prostaglandins such as PGIz and prostaglandin EZ (PGEa), and the attenu- ating effect of antihypertensive agents.

In keeping with this, Martin et al [17] reported that both indomethacin and ibuprofen significantly attenuated the pressor response to the mineralo- corticoids, fludrocortisone, and deoxycorticoster- one (DOCA). There was no increase in sodium re- tention during the period of indomethacin or ibu- profen administration. These findings support the evidence that prostaglandin inhibition plays a role

5A-44S May 17, 1991 The American Journal of Medicine Volume 90 (suppl 5A)

in the hypertensive action of sodium-retaining ste- roids.

RENAL PROSTAGLANDiNS AND BLOOD PRESSURE

It has become increasingly apparent over the years that the prostaglandins play a significant role in the modulation of blood pressure and in mediat- ing the effects of many antihypertensive agents. In the past, Lee et al [l&19] stressed the importance of the antihypertensive function of the kidney in the pathogenesis of hypertension. They proposed that some forms of hypertension are not solely the result of excessive activity of renal pressor mechanisms, but might also result from a deficiency in renal hy- potensive factors. The so-called renomedullary vas- odepressor substance, labeled medullin by Lee et al [B], subsequently was found to be a mixture of prostaglandins [19]. This deficiency might result in hypertension or might allow pressor systems to act unopposed to produce a hypertensive state.

The precise mechanism(s) by which the renal prostaglandins exert an antihypertensive effect is not clearly defined. Prostaglandins are ubiquitous substances that are produced in various tissues on demand and that exert their actions locally. Renal prostaglandin synthesis occurs in both the cortex and medulla. The major cortical prostaglandin is PGIz, which exerts an important positive influence on glomerular filtration rate. PGEz is the principal medullary prostaglandin.

An important renal effect of the prostaglandins is the promotion of natriuresis. Although there are several mechanisms by which renal prostaglandins promote natriuresis, PGEz directly contributes to this effect by inhibiting the tubular reabsorption of sodium and chloride. Inhibition of tubular respon- siveness to vasopressin (antidiuretic hormone) by PGEz favors free water excretion [20]. Therefore, NSAIDs, through prostaglandin (particularly PGEz) blockade may lead to sodium retention and edema, or, rarely, to an increased vasopressin ef- fect and hyponatremia. Sodium retention is a char- acteristic of virtually all NSAIDs, whether given alone or when added to diuretic therapy.

In studies [Sll of normal volunteers receiving ei- ther high- or low-sodium diets, agents such as aspi- rin, indomethacin, ibuprofen, and flurbiprofen have been reported to cause acute sodium retention, at least until an external balance with a fixed dietary intake can be achieved. Peripheral edema has been reported as a side effect in about 10% of patients receiving ibuprofen.

In addition to causing sodium retention when given alone, NSAIDs can attenuate the sodium ex- cretion caused by diuretics. Aspirin and indometha-

tin have been reported to attenuate the sodium ex- cretion produced by thiazides [22], loop diuretics [23], and spironolactone [24]. This action appears to be related to inhibition of intrarenal rather than extrarenal cyclo-oxygenase.

It has been suggested that sulindac is an NSAID that selectively spares intrarenal cycle-oxygenase [25]. This has been attributed to the fact that sulin- dac is a prodrug that is reduced in the body to its active sulfide metabolite [26], which in turn can be reoxidized and inactivated by mixed-function en- zymes in the kidney [27]. It has been reported [28] that sulindac does not suppress the urinary excre- tion of PGEz or 6-keto-prostaglandin, and that it does not cause sodium retention or attenuation of antihypertensive agents. The prostaglandin-spar- ing effects of sulindac, however, have been chal- lenged by others who report that, in doses of 400 mg/d, sulindac suppresses prostaglandin excretion [29-311 and causes sodium retention [29].

The prostaglandins have both an intrarenal and systemic vasodilatory effect that is an important mechanism through which antihypertensive effects can be produced. Nowak and Wennmalm [32] have examined changes in the regional blood flow of nor- mal humans during intravenous infusions of indo- methacin (0.3 mg/kg). The infusion of indomethacin produced a >‘75% reduction in urinary metabolites of prostaglandin E, a 30% increase in renal vascular resistance, and a 16% increase in splanehnic vascu- lar resistance. There was an accompanying rise in systemic mean arterial blood pressure of about 10 mm Hg, and a decrease in cardiac output of about 15% [331. Nonvisceral vascular resistance did not change significantly. These effects could be re- versed by the pharmacologic infusions of prosta- glandin El (4-8 pgimin).

Prostaglandins can also alter the vascular re- sponses to sympathetic nerve stimulation. PGEz or arachidonic acid infused into the renal artery of a rabbit can inhibit norepinephrine overflow caused by stimulation of the renal nerves [34]. This inhibi- tion could contribute ito the increased vascular re- sistance caused by NSAIDs in humans.

NSAlDs AND THE RENIN-ANiiIOTENSIN- ALDOSTERONE SYSTEM

There is evidence that NSAIDs may influence the renin-angiotensin-aldosterone system. Under normal circumstances, renin, an enzyme synthe- sized by the kidneys, is released into the circula- tion, where it acts on a plasma globulin substrate to form angiotensin I, a relatively inactive decapep- tide. Angiotensin I is then converted by angio- tensin-converting enzymes to angiotensin II, a po- tent vasoconstrictor and a stimulator of aldoste-

SYMPOSIUM ON CALCIUM ANTAGONISTS / HOUSTON

May 17, 1991 The American Journal of Medicine Volume 90 (suppl 5A) 5A-45S

SYMPOSIUM ON CALCIUM ANTAGONISTS I HOUSTON,

Arachidotiic Acid

cyclooxygenaee

Prostaglandins

i + + t 1 renal blood t Cl-(Na+) t ADH + renin

flow absorption in loop of Henle

I I + cjlomerular + water 4 aldosterone

iiltration rate

t proximai tubular Salt and Water 4 K+ absorption - Retention

) secretion

J Hyperkaierhia

Figure 3. Changes in renal functiori after NSAID blockade of cycle-oxygenase pathway and resul- tant inhibition of formation of the vasodilatoty prostaglandins. ADH = antidiuretic hormone.

rone. The ACE inhibitors exert their antihyperten- sive action. by blocking the conversion of angioten- sin I to angiotensin II. Angiotensin-converting en- zyme is identical to the enzyme that inactivates bradykinin. Therefore, ACE inhibitors, such as captopril, have the potential not only to antagonize the vasoconstrictor angiotensin II, but also to block the inactivation of bradykinin, a potent vasodilator. Bradykinin releases vasodilatory and natriuretic prostaglandins from renal and other tissues [35]. There is, therefore, a double mechanism whereby the NSAIDs block the antihypertensive effects of the ACE inhibitors. Several clinical studies support the antagonism of ACE inhibitors by NSAIDs [13- 161.

Evidence from studies [36,37] in animals in which angiotensin II was infused indicates that the pros- tanoids, particularly the vasodilatory PGIa and PGE2, counteract the vasoconstrictive activity of angiotensin II. Consequently, it is possible that an acute increase in renovascular resistance can be caused by the NSAIDs through their inhibition of vasodilatory prostaglandins stimulated by angio- tensin II and bradykinin.

Virtually all NSAIDs, including sulindac, have been associated with a decrease in plasma renin ac- tivity [4,11,13]. Prostaglandin-mediated stimuia-

tion of plasma renin activity by furosemide is blunted when individuals are pretreated with NSAIDs. Lopez-Ovejero et al [4] reported that in- domethacin reduced renin secretion by about 30% in untreated, uncomplicated hypertensive patients, and by about 75% in hypertensive patients receiv- ing therapy with diuretics or /3-adrenergic block- ers. This effect occurred in the absence of weight gain or an increase in arterial pressure in the un- treated hypertensive patients, suggesting that renin-suppressing effects of indomethacin can occur independently from the juxtaglomerular apparatus. NSAIDs have been associated with the combina- tion of hyporeninemia and hypoaldosteronism-a combination that may be responsible for the hyper- kalemia that is occasionally seen as a complication of NSAID therapy.

SUMMARY: NSAiDs AND ANTIi-h’PERiENSIVE DRUGS

The net effect of NSAIDs on antihypertensive drug therapy may be the sum of three actions, all of which involve prostaglandin inhibition. The marked reduction in plasma renin secretion acts to lower blood pressure. However, this effect is usually more than offset by salt and water retention from intrarenal prostaglandin inhibition and by the loss

5A-46S May 17, 1991 The American Journal of Medicine Volume 90 (suppl 5AJ

SYMPOSIUM ON CALCIUM ANTAGONISTS/HOUSTON

of the prostaglandins’ vasodilatory capabilities (Figure 3). NSAIDs appear more likely to attenu- ate antihypertensive drug therapy in patients with preexisting low-renin activity, i.e., the elderly and blacks, but their action is not limited to these groups.

Despite the wide range of antihypertensive agents whose effects are attenuated by NSAIDs, there are no reports of an interaction between NSAIDs and the calcium channel blockers, or cen- tral cx agonists. The calcium channel blockers have both a diuretic and a natriuretic effect and also de- crease systemic vascular resistance. The mecha- nisms by which they do so are apparently indepen- dent of the intrarenal and extrarenal prostaglandin system.

Our knowledge of the important interaction be- tween NSAIDs and antihypertensive agents is complicated by the many drugs in both categories. This complication has seriously limited the scope of the studies. There is an urgent need for more clini- cal studies involving large numbers of patients, more combinations of agents, and extending over prolonged periods. However, until these studies can be done, high-risk patients receiving NSAIDs should be identified, and blood pressure, renal func- tion, and serum potassium should be monitored fre- quently.

REFERENCES 1. Oates JA: Antagonism of antihypertensive drug therapy by nonsteroidal anti- inflammatory drugs. Hypertension 1988; ll(suppl II): 114-6. 2. Brown J, Dollery C, Valdes G: Interaction of nonsteroidal anti-inflammatory drugs with antihypertensive and diuretic agents: Control of vascular reactivity by endoge- nous prostanoids. Am J Med 1986; 8l(suppl 26): 43-57. 3. Wong DG, Lamki L, Spence JD, Freeman D, McDonald JWD: Effect of non-steroi- dal, anti-inflammatory drugs on control of hypertension by beta-blockers and diuret- ics. Lancet 1986; 1: 997-1001. 4. Lopez-Ovelero JA, Weber MA, Drayer JIM, Sealey JE, Laragh JH: Effects of rndo- methacin alone and durrng diuretic or beta-adreno-receptor blockade therapy on blood pressure and the renin system in essential hypertension. Clin Sci Mol Med 1978; 55: 203S-5s.

5. Watkins J, Abbott EC, Hensby CN, Webster J, Dollety CT: Attenuation of the hypotensive effect of propranolol and thiazide diuretics by indomethacin. Br Med J 1980; 281: 702-5. 6. Koopmans PP, Thien TH, Gribnau FWJ: Influence of non-steroidal anti-inflamma- tory drugs on diuretic treatment of mrld to moderate essential hypertension. Br Med J 1984; 289: 1492-4. 7. Patak RV, Mookerjee BK, Bentzel CJ, Hysert PE, Babej M, Lee JB: Antagonism of the effects of furosemide by rndomethacin in normal and hypertensrve man. Prosta- glandins 1975; 10: 649-59. 8. Durao V, Martins Prata M, Pires Goncalves LM: Modification of antihypertensive effect of beta-adrenoreceptor-blocking agents by inhibition of endogenous prosta- glandrn synthesis. Lancet 19n; 2: 1005-7. 9. Salvetti A, Pedrinelli R, Magagna A, Abdel-Haq B: The influence of indomethacin on some pharmacological acbons of atenoloi. In: Dinn MJ, Patron0 C, Cinotti G, eds. Prostaglandrns and the kidney. New York: Plenum Press, 1983; 287-95. 10. Salvetti A, Arzilli F, Pedrinellr R, Beggi P, Motolese A: Interaction between ox- prenolol and indomethacin on blood pressure in essential hypertensive patients. Eur J Clin Pharmacol 1982; 22: 197-201. 11. Ylitalo P, Pitkajarvi T, Pyykonen M-L, Nurmi A-K, Seppala E, Vapaatao H: Inhibi-

tron of prostaglandin synthesis by indomethacrn rnteracts with the antihypertensive effect of atenoloi. Clin Pharmacoi Ther 1985; 38: 443-9. 12. Rubin P, Jackson G, Blaschke T: Studres on the clinical pharmacology of prazo- sin. II: The influence of indomethacrn and of propranolol on the action and disposr- tion of prazosin. Br J Clin Pharmacoi 1980; 10: 33-9. 13. Salvetti A, Pedrinelli R, Magagna A, Ugenti P: Differential effects of selective and non-selective prostaglandin-synthesis inhibition on the pharmacological responses to captopril in patients with essentral hypertension. Clan SCI 1982; 63: 261S- 3s. 14. Silberbauer K, Stanek B, Templ H: Acute hypotensive effect of captopril in man modrfred by prostaglandin synthesis inhibition. Br J Clin Pharmacol 1982; 14: 87S- 93s. 15. Witzgall H, Scherer B, Weber PC: Involvement of prostaglandins in the actions of captopril. Clin Sci 1982; 63: 265S-7s. 16. Moore TJ, Crantz FR, Hollenberg NK, eta/; Contribution of prostaglandrns to the

anhhypertensive action of captopril in essential hypertension. Hypertension 1981; 3: 168-73. 17. Martin K, Zipser R, Horton R: Effect of prostaglandin inhibition on the hyperten- sive action of sodium-retaining steroids. Hypertension 1981; 3: 622-8. 18. Lee JB, Covino BG, Takman BH, Smith ER: Renomeduliary vasodepressor sub- stance, medullin: Isolahon, chemical characterrsation, and physiological properties. Clin Res 1965; 17: 57-U. 19. Lee JB: Prostaglandins and the renal antihypertensive and natriuretic endocrine function. Recent Prog Horm Res 1974; 30: 481-532. 20. Stillman MT, Schlesinger PA: Nonsteroidal anti-inflammatory drug nephrotoxic- ity, Should we be concerned? Arch Intern Med 1990; 150: 268-70. 21. Brater DC: Effect of indomethacin on salt and water homeostasis. Ckn Pharma- col Ther 1979; 25: 322-30. 22. Williams RL, Davies RO, Berman RS, eta/: Hydrochlorothiazide pharmacokinet- its and pharmacological effect: the influence of indomethacin. J Clin Pharmacol 1982; 22: 32-41. 23. Smith DE, Brater DC, Lin ET, Benet LZ: Attenuation of furosemide’s diuretic effect by indomethacin: Pharmacokinehc evaluation. J Pharmacokrnet Biopharm

1979; 7: 265-74. 24. Tweeddale MG, Ogilvie RI: Antagonism of spironolactone-induced natriuresrs by aspirin in man. N Engl J Med 1973; 289: 198-200. 25. Ciabattoni G, Crnotti GA, Pierucci A, et a/: Effects of sulindac and ibuprofen rn patients with chronic glomerular disease. N Engl J, Med 1984; 310: 279-83. 26. Duggan DE, Hare LE, Ditzler CA, Lei BW, Kwan KC: The disposition of suiindac. Clan Pharmacol Ther 1977; 21: 326-35. 27. Miller MJS, Bednar MM, McGiff JC: Renal metabolism of sulmdac, a novel non- steroidal antrrnfiammatory agent. Adv Prostaglandin Thromboxane Leukotriene Res 1983; 11: 487-91. 28. Sedor JR, Williams SL, Chremos AN, Johnson CL, Dunn MJ: Effects of sulindac and indomethacin on renal prostaglandrn synthesis. Clin Pharmacoi Ther 1984; 36:

85-91. 29. Roberts DG, Gerber JG, Barnes JS, Zerbe GO, Nies AS: Sulindac is not renal sparing in man. Clin Pharmacol Ther 1985; 38: 258-65. 30. Berg IU, Talseth T: Acute renal effects of sulindac and indomethacin in chronrc renal failure. Clin Pharmacol Ther 1985; 37: 447-52. 31. Brater DC, Anderson S, Baird B, Campbell WB: Sulindac does not spare the kidney. Clin Res 1983; 31: 868A. 32. NowakJ, Wennmalm A: Influence of indomethacin and prostaglandin E, on total and regional blood flow in man. Acta Physiol Stand 1978; 102: 484-91. 33. Wennmalm A: Influence of rndomethacin on the systemic and pulmonary vascu- lar resistance in man. Clin Sci 1978; 54: 141-5. 34. Malik KU, McGiff JC: Modulation by prostaglandins of adrenergic transmission rn the isolated perfused rabbit and rat krdney. Circ Res 1975; 35: 599-609. 35. Murthy VS, Waldrom TL, Goldburg ME: The mechanism of bradykinin potentia- tion after inhibition of angiotensin converting enzyme by SQ 14,255 in conscious rabbits. Crrc Res 1978; 43(suppl 1): 40-5. 36. Aiken JW, Vane JR: lntrarenal prostaglandin release attenuates the renal vaso- constrictor activity of angiotensin. J Pharmacol Exp Ther 1973; 184: 678-87. 37. McGiff JC, Crowshaw K, Terragno NA, Lonigro Al: Release of a prostaglandin-like substance into renal venous blood in response to angiotensin II. Circ Res 1970; 26/27(suppl I): 1-121-1-130. 38. Ryan JR, McMahon FG, Vargas R, et al: Differential Influences of salsalate, aspi- rin and naproxen on plasma renin activrty and platelet thromboxane synthesis, Arthritis Rheum 1986; 29(suppl): S103.

May 17, 1991 The American Journal of Medicine Volume 90 (suppl 5A) 5A-47S