997

prorenin normally functions as a myocardial growth factorand contributes to both physiological and pathologicalhypertrophy. This hypothesis would not only be consistentwith the degree of extraction reported here but might alsoexplain the cardiac growth of early pregnancy, when

maternal prorenin level is very high ;26,27 it may also providean explanation for the as yet often unexplained left

ventricular hypertrophy unrelated to afterload. 21 Perhapssecretion of prorenin by vascular resistance vessels,particularly renal vessels, and subsequent uptake by the heartmay constitute a circulatory control system influencingmyocardial contractility and regulating blood pressure andtissue perfusion. This work was supported by grants from the Australian Heart Foundation

(to S. L. S.), from the National Health and Medical Research Council ofAustralia (to J. D. H.), and from the William Buckland Foundation and theVictor Hurley Medical Research Fund (to J. A. W.).

Correspondence should be addressed to S. L. S.

REFERENCES

1 Leckie BJ. Inactive rerun: an attempt at a perspective. Clin Sci 1981; 60: 119-30.2 Sealey JE, Atlas SA, Laragh JH. Plasma prorenin: physiological and biochemical

characteristics. Clin Sci 1981; 63: 133s-45s.3. Atlas SA, Sealey JE, Hesson TE, et al. Biochemical similarity of partially inactive

renins from human plasma and kidney. Hypertension 1982; 4 (suppl II): 86-954. Bouknik J, Fehrentz JA, Galen FX, et al Immunologic identification of both plasma

and human renal inactive renin as prorenin. J Clin Endocr Metab 1985; 60:399-401

5. Takii Y, Figueiredo AFS, Inagami T. Application of immunochemical methods to theidentification and characterization of rat kidney inactive renin. Hypertension 1985;7: 236-43.

6. Imai T, Miyazaki H, Hirose S, et al. Cloning and sequence analysis of cDNA for humanrenin precursor. Proc Natl Acad Sci USA 1983; 80: 7405-09.

7 Lumbers ER. Activation of renin in human amniotic fluid by low pH. Enzymologia1971; 40: 329-36.

8. Kern MJ, Horowitz JD, Ganz P, et al. Attenuation of coronary vascular resistance byselective alpha1-adrenergic blockade in patients with coronary artery disease. J AmColl Cardiol 1985; 5: 840-46.

9. Thatcher R, Whitworth JA, Skinner SL. Changes in active and inactive renin withhaemodialysis. Nephron 1982; 32: 214-21.

10 Rapelli A, Glorioso N, Madeddu P, et al. Trypsin activation of active renin in humanplasma. As assessment of some methodological aspects. Clin Exp Hypertension 1981;3: 299-318.

11. Wahlqvist ML, Kaijser L, Lassers BW, Low H, Carlson LA. Glucocorticoid uptakeand release by the human heart: Studies at rest, during prolonged exercise, andduring nicotinic acid infusion. Scand J Clin Lab Invest 1972; 30: 261-66.

12 Keul J, Doll E, Steim H, Homburger H, Kern H, Reindell H. Die Substratversorgungdes gesunden menschlichen Herzens in Ruhe, wahrend und nach korperlicherAlbeit. Pflugers Arch ges Physiol 1965; 282: 1-27.

13 Esler M, Jennings G, Leonard P, et al. Contribution of individual organs to totalnoradrenaline release in humans. Acta Physiol Scand 1984; suppl 527: 11-16.

14. Nielsen MD, Giese J, Hesse B, Rasmussen S, Ibsen H. Inactive renin in renal venousblood: Biological, methodological and statistical aspects. Acta Med Scand 1983;suppl 677: 80-84.

15 Hesse B, Anderson ED, Ring-Larsen H. Hepatic elimination of renin in man. Clin SciMol Med 1978; 55: 377-82.

16 McKenzie IM, Reisin E, McKenzie JK. Uptake of inactive renin by human kidney.Clin Sci 1983; 65: 27-32

17 Webb DJ, Cumming AMM, Adams FC, et al. Changes in active and inactive renin andin angiotensin II across the kidney in essential hypertension and renal arterystenosis. J Hypertension 1984; 2: 605-14.

18 Huyduk K, Boucher R, Genest J. Renin activity content in various tissues of dogs underdifferent physiopathological states. Proc Soc Exp Biol Med 1970; 134: 252-55.

19 Re RN. Cellular biology of the renin-angiotensin systems. Arch Intern Med 1984; 144:2037-41.

20. Field LJ, McGowan RA, Dickinson DP, Gross KW. Tissue and gene specificity ofmouse renin expression. Hypertension 1984; 6: 597-603.

21 Derkz FHM, Wenting GJ, Man in’t Veld AJ, Verhoeven RP, Schalekamp MADH.Control of enzymatically inactive renin in man under various pathologicalconditions: Implications for the interpretation of renin measurements in peripheraland renal venous plasma. Clin Sci Mol Med 1978; 54: 529-38.

22. Johnson RL, Fleming NW, Poisner AM. Chromatographic and kinetic properties ofacid- and pepsin-activated inactive renin from human amniotic fluid. BiochemPharmacol 1979, 28: 2597-600.

23 Galen FX, Devaux C, Houot AM, et al. Renin biosynthesis by human tumoraljuxtaglomerular cells. Evidences for a renin precursor. J Clin Invest 1984; 73:1144-55.

24 Bowerman B, McKenzie IM, McKenzie JK. Inactive renin in peritoneal fluid-anindependent system or a transport form of circulating renin. J Hypertension 1984; 2:271-76.

25 Campbell DJ, Habener JS.The angiotensinogen gene is expressed and differentlyregulated in multiple tissues in the rat. J Clin Invest ( in press).

26. Skinner SL, Cran EJ, Gibson R, Taylor R, Walters WAW. Angiotensins I and II, activeand inactive renin, renin substrate, renin activity, and angiotensinase in humanliquor amnii and plasma Am J Obstet Gynecol 1975; 121: 626-30.

27. Hsueh WA, Leutscher JA, Carlson EJ, Grislis G, Fraze E, McHargue A. Changes inactive and inactive renin throughout pregnancy. J Clin Endocr Metab 1982; 54:1010-16.

28. Frohlich ED. Hemodynamics and other determinants in development of leftventricular hypertrophy Fed Proc 1983; 42: 2709-15

EFFECT OF NON-STEROIDALANTI-INFLAMMATORY DRUGS ON CONTROLOF HYPERTENSION BY BETA-BLOCKERS AND

DIURETICS

D. G. WONGL. LAMKI*

J. D. SPENCED. FREEMAN

J. W. D. MCDONALD

Departments of Medicine, Pharmacology and Toxicology, andNuclear Medicine, University of Western Ontario, London, Canada

Summary The effect of sulindac on renal function andblood pressure was compared with those of

placebo, piroxicam, and naproxen in 20 patients with

primary hypertension being treated with a diuretic and a beta-blocker. Although the three non-steroidal anti-inflammatorydrugs (NSAIDs) did not differ in their effect on renalfunction (weight, glomerular filtration rate, creatinine

clearance) or on serum thromboxane and plasma 6-ketoprostaglandin F1&agr; (6-keto PGF1&agr;), blood pressure was

significantly lower with sulindac than with placebo,piroxicam, or naproxen. These differences were associatedwith less renal cyclooxygenase inhibition by sulindac

(reflected by urinary thromboxane B2 and 6-keto PGF1&agr;) thanby other NSAIDs. The findings suggest that the bloodpressure differences reflect vasodilation due to differences inthe balance between systemic and renal effects of theNSAIDs.

Introduction

THE adverse effect of non-steroidal anti-inflammatorydrugs (NSAIDs) on renal function has been attributed partlyto inhibition of renal synthesis of vasodilator prostaglandinssuch as prostacyclin and prostaglandin E2. 1-12 NSAIDs arereported to cause fluid retention and hypertension, and mayimpair blood pressure control in hypertensive patientstreated with beta-blockers.3,13-22

Since sodium deprivation exacerbates the renal effects ofNSAIDs,6 these drugs might be expected to impair bloodpressure control in hypertensive patients being treated withdiuretics and beta-blockers. In our hypertension clinic

population of over 2000 referred patients, most of whom haveresistant hypertension, we have observed such an interactionin about 1% of patients annually.Because hypertension is common (it affects 30% of patients

over age 50 in our community),23 and because NSAIDs are incommon-use, 3 it seemed important to explore the interactionbetween NSAIDs and antihypertensive therapy, and to

determine whether any NSAID is less likely to affect bloodpressure control in patients on antihypertensive drugs.

Since there has been some evidence that sulindac, a

prodrug requiring metabolic conversion to its active sulphidemetabolite, is not excreted by the kidney in its active form24

*Present address: Department of Nuclear Medicine, M.D. AndersonHospital, Houston, Texas.

998

and has less effect on renal function than other NSAIDs,5,25,26we compared the effects of sulindac with those of placebo andtwo other commonly used NSAIDs, piroxicam and naproxen,on control of blood pressure in patients with primaryhypertension being treated with a diuretic and a beta-blocker.

Methods

20 patients (12 men, 8 women), aged 44-4±12.6 years, withprimary hypertension, kept stable for at least 6 months with beta-blockers and diuretics, agreed to participate in a study approved bythe University of Western Ontario committee on ethics in humanresearch. None of the patients had a blood pressure above 150 mmHg systolic or 100 mm Hg diastolic during the 6 months beforeentry, and none was on NSAIDs.

Subjects agreed to avoid all medication during the study exceptthose prescribed by the study team, and they were specificallycautioned about the existence of the large number of preparationscontaining acetylsalicylic acid. They were provided with a supplyof paracetamol tablets for relief of minor aches and pains that mightbe experienced during the study. In the event, too few tablets wereconsumed to make it worth analysing for differences betweentreatment groups. Patients with a history of peptic ulceration andpatients with serum creatinine levels outside the normal range forour laboratory were excluded, as were patients with any activemedical condition other than hypertension.The first phase of the study was a 4-week washout and titration

period in which patients attended weekly for adjustment of astandard antihypertensive - regimen. Patients were switched

abruptly from their previous diuretics and beta-blockers to a twice-daily regimen of timolol and a combination diuretic containinghydrochlorothiazide 50 mg and amiloride 5 mg, in an initial doseestimated to be equivalent to what they had been taking. Doses wereadjusted weekly to achieve control equivalent to or better than thepatient’s baseline. Patients then entered a randomised-sequencedouble-blind complete Latin-square crossover study, in which

antihypertensive treatment was left unchanged while patientsreceived 4 weeks of each of the following drugs-placebo 1 tablettwice daily, piroxicam 10 mg twice daily, naproxen 250 mg twicedaily, and sulindac 200 mg twice daily. There were no washoutperiods between the different regimens, but the design minimisescarry-over effect by ensuring that every drug is taken at some timeafter each of the other drugs. If carry-over were a severe problem,then differences between drugs would not be observed.Random allocation to drug sequences, provision of medication

bottles not bearing drugs names, and coding for drug identificationwere handled by the research pharmacist. Patients handed in theirmedication bottles to the pharmacy before attending the studycentre, to ensure that investigators remained blind to type ofNSAID. The code identifying the sequence of drugs received byeach subject was kept from the investigators until after the study,when all test results had been recorded.At the end of each month-long course of treatment, subjects

brought in a 24 h urine sample, a 50 ml portion of which was frozenfor subsequent measurement of6-keto PGFIa, the stable hydrolysisproduct of PGI2, and thromboxane B2 (TXB2), the stable

hydrolysis product of thromboxane A2. All were measured by theradioimmunoassay method of Ali and McDonald. Plasma 6-ketoPGFIa and serum thromboxane (reflecting platelet synthesis ofthromboxane during blood clotting ex vivo) were also measured,27as were creatinine clearance and 24 h urinary sodium excretion.Glomerular filtration rate (GFR) was measured b the use of99m-technetium diethylenetriaminepenta-acetic acid, 28and plasmarenin activity (PRA) was measured by radioimmunoassay,according to Cohen et al’s modification29 of the Haber method; thereagents were obtained from New England Nuclear, Canada, Ltd,Lachine, Quebec. Plasma renin samples were collected when thepatient arrived in the laboratory-ie, while ambulatory. Urinaryconcentrations of sulindac and of its sulphide and sulphonemetabolites were measured in all patients, and in a subgroup of 4patients, plasma and urinary naproxen levels, plasma piroxicam,and the urinary 5-hydroxy metabolite of piroxicam were measuredsemiquantitatively (ie, only an external standard was used)

by high performance liquid chromatography,30 using a

modification of the solvent strength to permit measurement of allthree NSAIDs.

Patients were weighed at each visit in the same state of undress,and supine blood pressure was taken as the mean of three readingsafter 5 minutes of recumbency, and standing blood pressure as themean of three readings after 2 minutes upright; measurements weremade with an automated digital oscillometric recorder (BPI 420,Medtek, Carollton, Texas), calibrated every 6 months against amercury manometer.

Analysis of variance was used to test for differences amongtreatments.

’

Results

All patients completed the study, with no importantadverse effects. Some patients had epigastric discomfort withsome of the regimens, but there were no significantdifferences between the groups.The four treatment groups did not differ in weight, urinary

sodium, serum creatinine, creatinine clearance, GFR, andPRA (table I).All three NSAIDs caused significant and equivalent

reduction of platelet thromboxane synthesis and of plasma6-keto PGF1G1’ concentration (fig 1). However, sulindac didnot differ from placebo with respect to renal synthesis ofprostaglandins, whereas both naproxen and piroxicamsignificantly reduced urinary TxB2 and 6-keto PGF1G1’ (fig 2).The treatment groups showed significant differences in

blood pressure control. Systolic blood pressures showed atrend (p=0’09) towards lower levels with sulindac and -

higher levels with piroxicam (fig 3). Both supine and standingdiastolic pressures, however, were significantly lower with

TABLE I-PRETREATMENT INDICES OF FLUID RETENTION AND RENAL

FUNCTION

Fig 1—Serum TxB2 (reflecting ex-vivo release of plateletthromboxane) and plasma 6-keto PGFIa (reflecting mainlypulmonary production of prostacyclin) in the four groups.

**p<O . 0 1; ***p<0 . 001.

999

Fig 2-Urinary TxB2 (reflecting mainly renal production ofthromboxane) and 6-keto PGF1&agr; (reflecting mainly renalproduction ofprostacyclin).

Fig 3-Supine and standing systolic blood pressure.*p<005. ,

sulindac than with placebo, and were significantly higherwith naproxen and piroxicam than with sulindac (fig 4). Thediastolic blood pressure changes were substantial-they wereabout the same as those which can be observed when anti-hypertensive drugs such as diuretics or beta adrenergicblockers are compared with placebo.Only trace amounts of the active sulphide metabolite of

sulindac could be detected in the urine; measurable amountswere found only in patients with high plasma levels ofsulindac (table II). The semiquantitative estimate of plasma

Fig 4-Supine and standing diastolic blood pressure.

*p<0-05; **p<0.01, ***p<0-001.

and urinary concentrations of all three NSAIDs and theirmetabolites in a subgroup of 4 patients (table III) puts theurinary levels of the sulindac metabolite in perspective.

Discussion

In 1961 Lee et al31 showed that extracts of renal medullainjected into rats treated with pentolinium and pentobarbitallowered blood pressure. They postulated that hypertension

TABLE II-PLASMA AND URINARY CONCENTRATIONS OF SULINDAC

AND ITS METABOLITE

TABLE III-SEMIQUANTITATIVE ESTIMATES OF-PLASMA AND URINARYCONCENTRATIONS OF THE THREE NSAIDS AND THEIR METABOLITES

1000

may not be due solely to an excess of factors that increaseblood pressure, but that it may result from a deficiency ofcompounds that lower blood pressure-in other words it

might be the result of an imbalance between vasoconstrictorand vasodilator influences. A deficiency of vasodilators mightthus produce hypertension even when levels of the pressorsubstances are normal. 32 Subsequently, Lee’s renal

hypotensive substance was shown to be a mixture of

prostaglandins.33Since then, renal prostaglandins have been shown to be

important in the regulation of blood pressure and in themodulation of the renin-angiotensin system.’-" The effects ofprostaglandins on renal blood flow and GFR, particularly insodium-depleted states,6 explains why NSAIDs, whichinhibit cyclooxygenase and thus the synthesis of renal

prostaglandins, have a major adverse effect on renal function,especially in patients with renal impairment. 3,26,34

NSAIDs also interact with antihypertensive drugsincluding beta-blockers, to reduce antihypertensiveefficacy.13-22 Although this interaction tends to be thought ofas one general to the entire class of NSAIDs,3 the work ofCiabattoni et al5,26 suggests that sulindac may differ from theother drugs in this respect. In patients with Bartter’s

syndrome and in patients treated with frusemide, doses ofsulindac large enough to produce 80% inhibition of

cyclooxygenase systemically had no renal effect. This

selective renal sparing action was in contrast to the effect ofindomethacin.5 Later, Ciabattoni et al showed that in patientswith chronic glomerular disease sulindac did not affect renalprostacyclin synthesis or renal function despite markedinhibition of extrarenal cyclooxygenase activity. In contrast,ibuprofen reduced urinary 6-keto PGF,, by 80%, decreasedcreatinine clearance by 28%, and increased serum creatinineby 40% after 1 week of treatment. 26These differences between sulindac and other NSAIDs are

thought to be due to differences in the disposition of the drug.The active form of sulindac is its sulphide metabolite. Thismetabolite is highly protein bound, has a long half-life, anddoes not appear in the urine in appreciable amounts. It isexcreted in the bile, whereas inactive metabolites are excretedin urine.24 Thus, to the extent that the renal prostaglandinsynthesis is affected by NSAIDs reabsorbed by the tubuleafter filtration, the kidney is probably not exposed to sulindacto the same extent as to other NSAIDs.

Interpretation of the piroxicam data is difficult.

Unpublished data (Twomey TM, personal communication)indicate that only 5&mdash;10% of piroxicam is excreted unchangedin urine. Our finding of only trace amounts of unchangedpiroxicam, taken together with the finding that the effect ofpiroxicam on urinary prostaglandins was equivalent to that ofnaproxen, suggests several possibilities-that an as yetunknown active metabolite of piroxicam is excreted by thekidney; that piroxicam is metabolised in the kidney to its5-OH metabolite (unlikely, since Twomey’s work35 indicatesthat the ratio of the glucuronide, which is almost certainlyhepatic in origin, to unconjugated 5-OH piroxicam is about3:1), that the small fraction of piroxicam that is excreted

unchanged is enough to exert the effect seen; that the 5-OHmetabolite may have unsuspected activity as a

cyclooxygenase inhibitor in renal tissue; or that thedifferences between the three NSAIDs are unrelated to the

pattern of excretion of active vs inactive forms of the drug.The low levels of 5-OH piroxicam reported in table in

represent the free metabolite; we did not measure the

glucuronide conjugate, which is thought to represent about75% of urinary 5-OH piroxicam (Twomey TM, personalcommunication).Although we expected that the effect of sulindac on blood

pressure control would be less than that of the other NSAIDs,we thought that the differences would be due to changes inrenal function and that fluid retention would be less withsulindac than with the other drugs. We were surprised,therefore, that the differences in blood pressure were notaccompanied by differences in renal function, and that bloodpressure was lower with sulindac than with placebo. Ourinability to detect differences between the NSAIDs in theireffect on renal function may have been due to inhibition ofrenin production by the beta blocker, since the activation ofthe renin-angiotensin system was less than expected in thepresence of diuretic treatment (table 1).There are few, if any, good examples of tissue-selective

cyclooxygenase inhibitors. Because its active metabolite isnot excreted by the kidney, sulindac may have a somewhatselective absence of inhibitory effect on renal cyclooxygenase.The differences in blood pressure effects may have been due

simply to weak cyclooxygenase inhibition by sulindac, but wethink that the serum thromboxane and plasma 6-keto PGF1alevels indicate that the doses of NSAIDs chosen were

equipotent with respect to their systemic (as opposed to renal)effects. Although we have no proof that the doses of the drugsused were equivalent in terms of their anti-inflammatorypotency, they did produce equivalent suppression of plateletcyclooxygenase activity, and they reduced plasma 6-ketoPGFIa to the same extent. Plasma 6-keto PGFIa is thought tobe derived largely from prostacyclin synthesis in the lung. Itseems therefore that sulindac, naproxen, and piroxicam, inthe doses used, produced equivalent reduction of

cyclooxygenase activity in at least two extrarenal tissues.It is possible that vasodilator prostaglandins such as PGI2

and PGE2 are actively synthesised in the kidney in thepresence of sulindac but not in the presence of other NSAIDs-in which case the deleterious effect of NSAIDs on blood

pressure control would be due to inhibition of synthesis ofvasodilator prostaglandins. The reduced inhibition of

urinary excretion of both TxB2 and of 6-keto PGFla inpatients taking sulindac showed that the drug spared renalcyclooxygenase activity. TxB2 is the hydrolysis product ofTxA2, a potent vasoconstrictor. The reduction in synthesis ofvasoconstrictors may have less effect than the reduction ofrenal synthesis of vasodilator prostanoids (PG 12 and PGE2)when diuretics are taken,6,36 as was the case in this study.Levenson et al6 have summarised evidence that PGE2synthesis is necessary for maintenance of renal cortical flow involume-depleted states, and the topic has been reviewedrecently by Carmichael and Shankel. 36

Although our findings suggest that sulindac is the NSAIDof choice for patients with hypertension, they should beextrapolated only with caution to other groups of patientsbecause our patients had normal renal function and theirrenin-angiotensin system was inhibited by beta-blockade.Two recent studies indicate that in patients with renal failure,the effects of sulindac are less deleterious effects than those ofother NSAIDs37,38 whereas Brater et al39 did not find anydifference between sulindac and other NSAIDs in normalvolunteers given frusemide. Recently, Puddey et al40 foundthat, in patients taking various antihypertensive drugs,sulindac did not impair blood pressure control to the sameextent as did indomethacin; they also showed a discrepancy

1001

between the effects of indomethacin and sulindac on urinaryprostaglandins that would tend to support our findings.Further study in hypertensive patients with impaired refialfunction, and on other antihypertensive regimens, is

required.We thank the patients for their participation, and Mrs Pat Manso for her

coordination of the data collection. D. G. W. was a recipient of a researchfellowship funded by Merck Frosst Canada Ltd and awarded by the CanadianSociety for Clinical Pharmacology.

Correspondence should be addressed to J. D. S., Hypertension Unit,Department of Medicine, Victoria Hospital, London, Ontario, Canada N6A4G5.

REFERENCES

1. McKay IG, Muir AL, Watson ML. Contribution of prostaglandins to the systemic andrenal vascular response to frusemide in normal man. Br J Clin Pharmacol 1984; 17:513-19.

2 Lavelle KJ, AronoffGA, Luft FC The effect of nonsteroidal anti-inflammatory agentson renal function. J Indiana State Med Assoc 1982; 75: 334-36.

3. Clive DM, Staff JS. Renal syndromes associated with nonsteroidal anti-inflammatorydrugs N Engl J Med 1984; 310: 563-72.

4 Zawanda ET. Renal consequences of nonsteroidal antiinflammatory drugs PostgradMed J 1982; 71: 223-30.

5 Ciabattoni G, Pugliese F, Cinotti GA, Patrono C. Europ J Rheum Inflamm 1980; 3:210-21.

6. Levenson DJ, Simmons CE, Brenner BM. Arachidonic Acid metabolism,prostaglandins, and the kidney Am J Med 1982; 72: 354-57.

7 Dunn MJ, Zambraski EJ. Renal effects of drugs that inhibit prostaglandin synthesis.Kid Int 1980; 18: 609-22

8. Halushka PV, Wolhtmann H, Privitera PJ, Hurwitz G, Margolus HS. Bartter’ssyndrome; prostaglandins and indomethacin effects. Ann Intern Med 1977; 87:281-86.

9 Horton R, Zipser R, Fichman M. Prostaglandins, renal function and vascular

regulation. Med Clin N Am 1981; 65: 891-91410. Terragno NA, Malik KU, Nasjletti A, Terragno DA, McGiff JC. Renal

prostaglandins. Adv Prostagl Thrombox Res 1976; 2: 561-71.11. Weber PC, Siess W. Interactions of renal prostaglandins with the renin-angiotensin

system. Pharmac Ther 1982; 15: 321-37.12. Weber PC, Scherer B, Siess W, Held E, Schnermann J Formation and action of

prostaglandins in the kidney. Klin Woch 1979; 57: 1021-29.13 Mills EH, Whitworth JA, Andrews J, Kincaid-Smith P. Non-steroidal anti-

inflammatory drugs and blood pressure. Aust NZ J Med 1982; 12: 478-82.14. Sullivan JM Prostaglandins and regulation of blood pressure: clinical implications.

Pharmac Ther 1982; 15: 447-65.15. Heinemann HD, Lee JB Prostaglandins and blood pressure control. Am J Med 1976;

61: 681-95.

16. Durao V, Prata MM, Goncalves LMP. Modification of antihypertensive effect of betaadrenoceptor-blocking agent by inhibitor of endogenous prostaglandin synthesis.Lancet 1977; ii 1005-07

17 Salvetti A, Arzilli F, Pedrinetti R, Beggi P. Interaction between oxprenolol andindomethacin on blood pressure in essential hypertensive patients. Eur J ClinPharmacol 1982; 22: 197-201.

18 Watkins J, Abbott EC, Hensby CN, Webster J, Dollery CT. Attenuation ofhypotensive effect of propranolol and thiazide diuretics by indomethacin. Br Med J1980; 281: 702-05

19 Gerber JG. Indomethacin-induced rises in blood pressure Ann Intern Med 1983; 99:555-58.

20 Frohlich JC, Rosenkranz B. Role of prostaglandins in the regulation of blood pressure.Cardiovascular pharmacology of the prostaglandins. New York: Raven Press, 1982:259-65.

21 Lee JB. The renal prostaglandins and blood pressure regulation Adv ProstaglThrombox Res 1976; 2: 573-85.

22 Lee JB, McGiff JC, Kannegiesser H, Aykent YY, Mudd JG, Frawley TF.Prostaglandin A: antihypertensive and renal effects. Ann Intern Med 1971; 74:703-10

23 Bass MJ, Donner A, McWhinney IR. Effectiveness of family physicians in

hypertension screening and management. Can Fam Phys 1982; 28: 225-58.24 Duggan DE, Hare LE, Ditzler CA, Lei BW. The disposition of sulindac. Clin

Pharmacol Ther 1977; 21: 326-35.25 Bunning RD, Barth WF. Sulindac: a potentially renal sparing nonsteroidal

antiinflammatory drug. JAMA 1982; 248: 2864-67.26 Clabattoni G, Cinotti GA, Pierucci A, Simonetti BM, Patrono C. Effects of sulindac

and ibuprofen in patients with chronic glomerular disease. N Engl J Med 1984,310:279-83.

27 Ali M, McDonald JWD. Synthesis of thromboxane B2 and 6-keto-prostaglandin F1-alpha by bovine gastric mucosa and muscle microsomes. Prostaglandins 1980; 20:245-54

28 Lamki L, Spence JD, MacDonald AC, Roulston M Differential glomerular filtrationrate (GFR) in diagnosis of renovascular hypertension and follow up of balloonangioplasty. Clin Nucl Med (in press).

29 Cohen EL, Grim CE, Conn GW, et al. Accurate and rapid measurement of plasma reninactivity by radioimmunoassay. J Lab Clin Med 1971; 77: 1025-38.

30 Dusci LJ, Hackett LP. Determination of sulindac and its metabolites in serum by highperformance liquid chromatography. J Chromotography 1979; 171: 490-93.

31 Lee JB, Hickler RB, Saravis CA, Thorn GW Sustained depressor effects of

renomedullary extracts in the normotensive rat Circulation 1962; 26: 747-48.

32. Lee JB. Prostaglandins and the renal antihypertensive and natriuretic funtion Rec ProgHorm Res 1974; 30: 481-532.

33. Lee JB, Covino BG, Takman BH, Smith ER. Renomedullary vasodepressor substance,medullin: Isolation, chemical characterisation and physiological properties. CircRes 1965; 17: 57-77.

34 Linton AL Adverse effects of NSAIDs on renal function. Can Med Ass J 1984; 131:189-91.

35. Hobbs DC, Twomey TM. Biotransformation of piroxicam by man. Fed Proc 1978; 37:271(abstract).

36. Carmichael J, Shankel SJ. Effects of Nonsteroidal anti-inflammatory drugs onprostaglandins and renal function. Am J Med 1985; 78: 992-1000

37. Swainson CP, Griffiths P. Acute and chronic effects of sulindac on renal function inchronic renal disease. Clin Pharmacol Ther 1985; 37: 298-300.

38. Berg KJ, Talseth T. Acute renal effects of sulindac and indomethacin in chronic renalfailure. Clin Pharmacol Ther 1985: 37: 447-52.

39. Brater DC, Anderson S, Baird B, Campbell WB. Effects of ibuprofen, naproxen, andsulindac on prostaglandins in men. Kidney Int 1985; 27: 66-73

40. Puddey IB, Beilin LJ, Vandongen R, Banks R, Rouse I. Differential effects of sulindacand indomethacin on blood pressure in treated essential hypertensive subjects ClinSci 1985; 69: 327-36.

EFFECT OF ACETAZOLAMIDE ON EXERCISEPERFORMANCE AND MUSCLE MASS AT HIGH

ALTITUDE

A. R. BRADWELL

J. H. COOTEJ. J. MILLES

P. W. DYKESP. J. E. FORSTER

I. CHESNERN. V. RICHARDSON

AND BIRMINGHAM MEDICAL RESEARCH EXPEDITIONARYSOCIETY

University Departments of Immunology, Medicine, and Physiology,Medical School, Birmingham University, Birmingham

Summary The effect of acetazolamide (Az) on exerciseperformance and muscle mass in

acclimatised subjects at an altitude of 4846 m was assessed in11 subjects and compared with the effect of placebo on 10other subjects. Exercise performance at 85% maximum heartrate fell by 37% in the Az group and by 45% in controls(p<0&middot;05). Weight loss was greater in the placebo group athigh altitude (p<0&middot;01) and this correlated with the fall inexercise performance (p<0&middot;001). During the expeditionanterior quadriceps muscle thickness fell by 12&middot;9% in thecontrol group and 8&middot;5% in the Az group (p<0&middot;001), while

biceps muscle thickness fell by 8&middot;6% in controls and 2&middot;3% inthe Az group (p<0&middot;001). Measurements of skin-foldthickness indicated a loss of 18% of total body fat in theplacebo group and 5% in the Az group by the end of theexpedition (p<0&middot;001). Calorie intakes at altitudes above3000 m were low and similar for the two groups. The Az

group had fewer symptoms of acute mountain sickness butdifferences between the two groups were not statisticallysignificant. Acetazolamide is therefore useful for climbersand trekkers who are acclimatised to high altitudes. It couldbe most useful at extreme altitudes, where maintenance ofexercise performance and muscle mass are important.

Introduction

ACETAZOLAMIDE (Az) is useful in the prophylaxis of acutemountain sickness (AMS).1-3 By stimulating ventilation itraises alveolar oxygen levels and thus increases the amount of

oxygen that can be taken up by the blood. The effect ofAz hasbeen examined in trekkers going up to altitudes of 5000 m butits use by climbers has been limited for several reasons. First,there are no data on its use by acclimatised subjects withoutAMS (the majority of climbers). Second, Az reduces exerciseperformance at sea level,4 which could have serious

consequences if it occurred at altitude. Third, there are no