Plasma noradrenaline levels ot 100 W

(pg mL-1l

Figure3 Lack of correlation between exercise-induced stimulation of plasma noradrenaline

and antihypertensive effects of chronic treatment with propranolol. (From Bühler et al., 1980;

withpermission).

In arecent investigation (Bühler & Kiowski, 1980), a relationship between

pretreatment (100 W) exercise levels of noradrenaline and the hypotensive response

to propranolol in 21 subjects (Figure 3) and other ß-adrenergic blocking drugs in 16

further observations could not be demonstrated. This fmding notably differed from

the earlier observation by DistIer, Keim, Cordes, Philipp & Wolff (1978). Sixteen

essential hypertensive subjects were inve stigated and the increase in plasma

noradrenaline produced by ergometer bicycling was used as an index of the

responsiveness of the sympathetic nervous system. Patients were then treated with

atenolol, 100 mg twice a day for several weeks. The initial plasma noradrenaline

concentration during exercise and the subsequent hypotensive response to atenolol

were significantly correlated (r = 0.63). The correlation coefficient increased to 0.84

when plasma noradrenaline during exercise was related to baseline concentration

(Figure 4). It is of note, that systolic pressure elevation during exerise was related to

the rise in noradrenaline and was also a fair predictor of the response to

ß-adrenoceptor blockade.

Watson, Erikson, Hamilton, Reid, Stallard & Littler (1980) studied 17 essential

hypertensive patients. Sampies for measurements of noradrenaline and adrenaline

were taken at rest and after bicycle ergometry. The hypotensive response to

propranolol (240 mg daily) and other ß -adrenergic receptor blocking drugs, was

found to be unrelated to pretreatment levels of adrenaline or noradrenaline at rest

and during exercise and to the ratio between those levels. The difference between the

results ofDistier et al. (1978) on the one hand, and Bühler et al. (1977) and Watson et

al. (1980) on the other is difficult to explain. Apart from such imponderables as

patient selection and assay technique one could venture the explanation, that the

physical challenge put to use by Distier et al. (1978) was much more sthenie (200 W

during 2 min) than the challenge in the other studies (which tended to be more

260 W. H. BIRKENH ÄGER & P. W. OE LEEUW

gradual and thus eould have failed to elieit sympathetie nervous reaetivity to its full

extent). Plasma noradrenaline during exercise

(ng litre- I

)

o 1000 2000

0-+--------->--------'----

..

.. . 10

20

e::J

.,

., 30

e y a 0 009 l' 6 .4 4 Co

"0 r - 0 .632

0 P cO.Ol 0 40 z

ö

.;: Plasma noradrenaline .. (exercise Ibaseline)

;; 0 5 10 15 20

c 0 0

..

E

-0 10

...,

0

e

0

.. 20

0

30

Y -2.04l'1.50

r - 0.640 40 P -e 0.001

A

B

Flgure 4 Correlation between parameters of sympathetic responsiveness and subsequent

decrease in mean arterial pressure following treatment with atenolol. A. Plasma noradrenaline

concentration at the end of bicycle exercise. B. Ratio of exercise to baseline plasma

noradrenaline concentration. (Frorn DistIer et al., 1978; with perrnission).

Reeently, different types ofadrenergie stimulation were examined under metabolie

ward eonditions (de Leeuw, Wester, Willemse & Birkenhäger, 1980). One stress test

eonsisted of 3-5 min of isometrie exereise (30% of maximal handgrip force). In the

other challenge study, the subjeet was exposed, during 3-5 min, to the risk of an

exerueiating noise when he failed to pass a dexterity test. Blood sampies for

noradrenaline assa y were taken at rest and after 3 and 5 min of stimulation. One

group of25 patients was then treated with atenolol (average dose 150 mg daily). Both

noradrenaline levels at rest and those obtained during stimulation failed to prediet

the hypotensive response.

However, when the same proeedure was applied to 15 patients who were treated

with prazosin, a highly significant relation was found between noradrenaline

inerements after 3 min of stimulation and the drug-indueed deerease in blood

pressure (r = 0.80 , Figure 5). The inerease in blood pre ssure during stimulation

showed a mueh weaker relationship with the hypotensive effeet ofprazosin.

Other types of treatment

Hong Tai Eng, Huber-Smith & MeCann (1980) investigated the sensitivity of 20

(normal renin) hypertensive patients to sodium restrietion (and subsequent diuretie

treatment). Blood for eateeholamine assay was eollected after 20 min reeumbeney; a

CAT ECHOLAMINE LEVELS AN D ANT IHYPERTENSIVE TH ERAP Y 261

second sampie was drawn after 5 min ofstanding. Both were evaluated with a view to

the hypotensive response to sodium restriction from approximately 140 to 70 mmol

daily. Mean arterial pre ssure fell in 12 subjects. The overall correlation between the

change in blood pressure and the change in sod ium excretion was not significant. The

decrease in mean arterial pre ssure for II subjects with normal plasma noradrenaline

« 374 pg mi-I) was 10.8 ± 10.6 mm Hg, while for those with high plasma

noradrenaline (> 374 pg ml'") it was 6.8 ± 9.1 mm Hg. Although th is fmding

roughly reflects the expected trend, according to the volume-vasoconstrictor analysis,

it ob viously fails to make a strong case for catecholamine profiling.

10 •

•

• • •

• •

• •

•

•

•

•

5 •

e::> 0 • '"

'"

'"a

:g

-5

<;

c:

0

'"E

c - 10

'"

'"

c:

0

.c

U - 15

- 20

I I I I

0 .04 0.08 0 .12 0.16 0. 20

Stimulated noradrenoline- B0501 norodrenoline

(ng ml-I ) (untreoted)

Figure 5 Relation ship between increases of plasma noradrenaline during 3 min static exercise

and subsequent hypoten sive response to treatment with prazosin. (From de Leeuw et al., 1980).

Discussion

The concept that adrenergic tone (as estimated by measuring one or several pla sma

catecholamines) is a determinant of responsiveness to anti-adrenergic (mainly ßadrenoceptor blocking-) drugs in essential hypertensive patients has been tested. It is

obvious from Table I that a simple relationship is lacking. Attempts have been made

to safeguard the principle of 'catecholamine profiling' as a predictive index by the

addition ofsome refmements and arabesques to the procedure. Esler et al. (1977), by

introducing an element of dose responsiveness, were indeed able to identify a

subgroup ofpatients with elevated noradrenaline levels and enhanced responsiveness

to low levels ofpropranolol. This fmding would seem to indicate that catecholamine

profiling would serve a practical purpose; howe ver no clear improvement was

obtained in comparison with the results of renin profiling in thi s study. A further

attempt to establish a role for catecholamine profiling in the approach to essentia l

hypertension consisted of testing the reactivity rather than baseline acti vity of the

sympathetic nervou s system. Th is has been achieved by means of a number of

provocative procedures: ergometric or isometric exercise, gravitation al or mental

stress. Table 1again reveals that fmdings are inconsistent.

262 w. H. BIRKENHÄGER & P. W. DE LEEUW

Table I Overview of data relating parameters of adrenergie (re)aetivity to subsequent

hypotensiveeffeetiveness of ß-and/or a-adrenergie reeeptor blockingagents.

Relationship

0/adrenergic

Number Condition activityto

0/ 0/ bloodpressure

Reference subjects sampling Treatment reduction

Birkenhägeret al. (1977) 26 Reeumbeney Propranolol Absent

Esleret al. (1977) 23 Reeumbeney Propranolol Present{low dosage)

Absent(high dosage)

Muiesanet al. (1978) 14 Reeumbeney Labetolol(i.v.) Present

DistIeret al. (1978) 16 Dynamieexereise Atenolol Present

v. resting

Bühleret al. (1980) 21 Dynamieexereise Propranolol Absent

7 Dynamieexereise Atenolol Absent

9 Dynamieexereise Pindolol Absent

Watson et al. (1980) 17 Dynamieexereise Propranolol Absent

v. resting &other

-adrenoceptor

lockingdrugs

De Leeuwet al. (1980) 25 Isometrieexereise Atenolol Absent

v. resting

(1980) 15 Isometrieexereise Prazosin Present

v. resting

It is not clear why such rather equivalent investigations as the one by DistIer et al.

(1978) on the one hand and those by Bühler & Kiowski (1980) and Watson et al.

(1980) on the other have yielded opposite results. If patient selection was a crucial

factor, then that fact alone would condemn the principle, since profiling itself is

intended to be a base for patient selection. The best explanation at hand would be,

that DistIer et al. (1978) applied a much more acute and demanding stimulus than the

other groups. Assuming that this is true, many clinicians would be inclined to detract

this from the value of the method of 'catecholamine profiling', since exercising to

near exhaustion would seem to be rather prohibitive from a clinical point ofview.

Although the predictive value ofnoradrenaline increase during static exercise with

respect to treatment with prazosin (though not with ß-adrenergic receptor blocking

drugs) was impressive (de Leeuw et al., 1980) it is very much doubted whether this is

of any other than theoretical significance. The intricate interrelations between the

different mechanisms ofblood press ure regulation would require the implementation

of a very accurate protocol, including dietary and environmental precautions, before

the actual assessment of thc catecholamine profile could bc carried out. This

approach - even if it would be productive on a small scale - would bc basically

conflicting with the need to take care of a multitude of hypertensive patients. On

balance, therefore, the adoption ofa time-consuming procedure, like the present one,

for thc 'individualization' ofantihypertensive treatment is not advocated.

For the near future, it is to be expected that only obvious and easily obtainable

factors such as blood pressure itself, heart rate and othcr parameters oftenseness, age,

sex, race and profession will be the weights to load the dice with, when a therapeutic

decision is made.

CATECHOLAMINE LEVELSAND ANTIHYP ERTENSIVE THERAPY 263

Conclusions

In the past, many attempts have been made to rationalize the choice of antihypertensive drug in an individua l patient. The earlier parameters (haemodynamic

indices, renin/sodium index) were supposed to reflect mainly the degree of

sympathetic nervous system activity. It was therefore to be expected that attention

would move to circulating catecholamines as a more direct expression of adrenergic

activity. Despite various precautions, catecholamine (mainly noradrenaline)

profiling a ppea rs to be a rather unreliable method of predicting the hypotensive

response to ß-adrenoceptor blockade. In addition, the implementation of this

method would demand too much effort and patience on the part ofboth patient and

doctor to serve a practical purpose.

References

Birkenhäger, W. H., de Leeuw, P. W., Kho, T. L., Wester, A., Vandongen, R. & Falke, H. E.

(1977). Selection of hyperten sive patients for treatment with betablockers. In Betablockade, ed. Mänzer, W., Schömig, A., Dietz, R. & Liehtlen , P. R. pp. 113-121. Stuttgart:

G. Thieme Verlag.

Bühler, F. R., Burkart, F., Lütold , B. D., Küng, M., Marbet, G. & Pfisterer, M. (1975). Antihypertensive betabloeking action as related to renin and age: a pharmaeological tool to

identify pathogenetie meehanisms in essential hypertension. Am. J. Cardiol.. 36, 653.

Bühler, F. R., Burkart, F., Lütold , B. E., Bertel, O. & Pfisterer, M. (1977). Plasmakateeholamine und Hämod ynamik im Verlauf der antihvpertensiven Betablockade:

verschiedene Muster bei " Propranolol-Responders" und "Non-Responders". Schweiz.

Med. Wochenschr., 107, 1590-1591.

Bühler, F. R. & Kiowski, W. (1980). Plasma eateeholamines, renin , age and antihypertensive

response to betablockers . In Proceedings 01 the Se venth International Congress 01

Pharmacology, ed. Vanhoutte, P. pp. 322-328. New York: Raven Press.

de Leeuw, P. W., Wester, A , Willemse, P. J. & Birkenh äger. W. H. (1980).Elfeets of'prazosin on

plasma noradrenaline and plasma renin coneentration in hyperten sive subjects. J.

Cardiovasc. Pharmac., in press.

Distler, A, Keim, H. J., Cordes, U., Philipp, T. & Wollf, H. P. (1978). Sympathetie

responsiveness and antihypertensive elfeet of beta reeeptor blockade in essential

hyperten sion. Am. J. Med., 64,446-451.

Esler, M. D. & Nestei, P. J. (1973). Renin and sympathetie nervou s system responsiveness to

adrenergic stimuli in essential hypertension. Am. J. Cardiol., 32,643-649.

Esler, M., Zweifler, A., RandalI , O. & Dequattro, V. (1977). Pathophysiologie and pharmaeokinetic determinants of the antih yperten sive response to propranoloI. Clin. Pharmac.

Th ,22 299-308

Esler. M. D., Zweifler. A., Randall. 0 ., Juliu s, S. & Dequattro, V. (1 978). The determinants of

plasma-renin activity in essential hypertension. Ann. intern. Med.. 88, 746-752.

Fournier, A., Hardin, J. M.. Alexandre , J. M., Lombaert , M., Ronco, G., Bezoe, J. F.• Desmet.

G. & Quichaud , J. (19 76). Antihypertensive elfect of acebutolol: its relation to sympathetic

nervous system responsiveness and to plasma renin and doparnine-jj-hydro xylase

activities. Clin. Sci. Mol. Med., 51, 477-480( s).

Hong Tai Eng, F. W., Huber-Smith , W. & McCann , D. S. (1980). The role of sympathetie

aetivity in normal renin essential hypertension. Hypertension, 2, 14.

Muiesan, G., Agabiti-Rosei, E., Alieandri , C. & Fariello , R. (1978). Sympathetie aetivity and

haemodynamie ehanges after eombin ed alpha- and betablockade in hypertension. In

Abstract Book VIII World Congress 01 Cardiology, Abs. no. 1028, p.348, 17-23 Sept.

1978,Tokyo, Japan . International Soeiety and Federat ion of Cardiology.

Watson, R. D. S., Erikson, B. M., Ham ilton, C. A , Reid, J. L., Stallard, T. J. & Littler , W. A.

(1980). Elfects of ehron ie beta-adrenoeeptor antagonism on plasma eateeholamines and

blood pressure in hypertension. J. Cardiovasc. Pharmac., in press.

ß-ADRENOCEPTOR BLOCKING DRUGS

ANDTHE

RENIN-ANGIOTENSIN SYSTEM

J. MENARD, P. BAUTIER, P. F. PLOUIN

M. THIBONNIER & P. CORVOL

I NSERlvf Unite 36,

17rue du Fa aMoulin,

Paris 75005, France

Biological tests of the hypertension work-up are used for two different purposes: to

determine diagnosis and treatment and to detect rare diseases, For all patients, it is

useful to conduct initial tests of plasma creatinine and potassium, and, very often , of

plasma glucose, cholesterol and uric acid. A second group of tests is useful in a few

patients for detection ofsurgically curable causes of hypertension .

As regards plasma renin activity (PRA) measurements, these would fit into the first

group of tests if they were demonstrated to constitute a completely reliable basis for

the choice of therapy. Some investigators have reported that patients with low PRA

are very responsive to diuretics (Vaughan, Laragh, Gavras, Bühler, Gavras, Brunner

& Baer ,1973), and patients with normal or high PRA are more responsive to

ß-adrenoceptor blocking drugs (Bühler, Laragh, Baer , Vaughan & Brunner, 1972).

These observations form the basis of the vasoconstriction analysis for the understanding and treatment of hypertension (Laragh, 1976). They imply that low renin

hypertension is the result of a more or less positive sodium balance, and that normal

and high renin hypertension are partly or completely dependent on the reninangiotensin system. According to this view, routine measurement ofPRA might help

to improve the choice of the initial monotherapy in hypertension . Moreover the

decrease in ren in secretion induced by ß-adrenoceptor blocking drugs would participate in their antihypertensive effect , whereas an excessive renin response to diuretic

therapy would limit the antihypertensive potential of this treatment. Nevertheless,

the fmdings of Bühler et al. (1972) have not been confirmed in some reports

(Birkenhäger , de Leeuw, Wester, Kho, Vandongen & Falke, 1977) and the

mechanism of the antihypertensive etTects of ß-adrenoceptor blocking drugs is still

controversial.

In this report, the research fmdings of this group, from 1974 to 1980 will be

reviewed, They consistently confirrn that blood pre ssure fall induced by

ß-adrenoceptor blocking drugs is significantly correlated to initial PRA. An atternpt

to explain the divergent results obtained in the literature will be made and the

practical and theoreticallimitations ofthese fmdings will be discussed.

RENIN AND ß -ADRENOCEPTOR BLOCKING DRUGS 265

0.001 Ducrocq et al. (1975)

0.001 Menard et al.(1976)

0.001 Alhenc-Gelas et al.(1978)

0.05 Plouin et al.(1979)

0.001 Kreft et al.(1979)

0.001 Thibonnier et al.(1980)

0.001 Plouin et al. (1980)

Correlations between PRA and the blood pressure response to ß-adrenoceptor

blocking drugs

The results ofseven successive trials are summarized in Table I (Ducrocq, Degoulet,

Charpentier, Corvol & Menard, 1975; Menard, Bertagna, N'Guyene, Degoulet &

Corvol, 1976; Alhenc-Gelas, Plouin, Ducrocq, Corvol & Menard, 1978; Plouin,

Comoy, Bohuon, Corvol & Menard, 1979; Kreft, Menard & Corvol, 1979;

Thibonnier, Lardoux & Corvol, 1980; Plouin, Alhenc-Gelas, Degoulet, Corvol &

Menard, 1980). In each study, PRA was measured by radioimmunoassay of angiotensin I, after one hour of ambulation (08.00-09.00 h), on the third day in hospital,

with normal sodium intake (100 mmol daily). Patients were given no antihypertensive drugs for 15 days before the investigation. All seven studies included

subjects with low, normal and high renin levels . (Normal values: 0.7 - 3 ng mi-I h-I).

The value of renin measurements in predicting the antihypertensive efTects of

ß-adrenoceptor blocking drugs was estimated in six hospital trials, after two days of

treatment, and one conducted in the outpatients' clinic for six weeks (Alhenc-Gelas

et al., 1978). The very significant correlation observed between the fall in blood

pressure measured on the third day of acebutolol administration and after six weeks

oftreatment has also been reported for other ß-adrenoceptor blocking drugs (Bühler,

Burkart, Lutold, Kung, Marbet & Pfisterer, 1975; Lehtonen, 1976; Collste & Haglund, 1979). No correlation was found between PRA and the fall in blood pressure

within any of the subgroups of patients with low, normal and high renin levcls

respectively.

Although highly significant, these correlations betwecn PRA and the decrease in

blood pressure aftcr ß-adrenoceptor blockade do not account for a sufficient

variance to allow prediction of the efficacy of ß-adrenoceptor blocking drug

treatment in the individual patient. Besides the cost and technical difficulties

involved in obtaining accurate measurements of renin from both the clinical and

laboratory viewpoints, these tluctuations from one individual to another greatly limit

the practical interest of these correlations. Consequently, routine measurements of

PRA are at present restricted to patients referred to hypertension clinics for special

reasons.

Table 1 Correlation between initial plasma renin activity (PRA) and the fall in blood pressure

induced by ß-adrenoceptor blockade in seven successive studies.

Number

Study of P Reference

subjects

Systolic time intervals and

antihypertensive effect of

acebutolol. 31 0.580

Inpatient evaluation of

acebutolol. 44 0.498

Outpatient evaluation of

acebutolol. 30 0.639

Plasma noradrenaline and the

antihypertensive effect of

acebutolol. 9 0.683

Saralasin test and

acebutolol treatment. 24 0.620

Comparative trial of

labetalol and acebutolol. 16 0.739

Stepped-care treatment of

hypertension 60 0.520

r =correlation coefficient

p= indicates the level of'significance ofthe correlation

266 J. MENARD, P. BAUTIER, P. F. PLOUIN, M. THIBONNIER & P. CORVOL

Tentative explanations for the lack of correlation between PRA and the blood

pressure response to ß-adrenoceptor blocking drugs

It is hardly conceivable that so many divergent results may be recorded in the

medical literature for a correlation between two parameters wh ich are apparently

easy to measure.

The lack of predictive value of renin measurement is easy to explain in many

studies by such factors as the acidification of plasma and the measurement of total

renin (Meekers, Missotten, Fagard, Demuynck, Harvengt, Pas, Billiet & Arnery,

1975), the absence ofstimulation of renin secretion in the upright posture (Hansson ,

1973), the failure to include patients with high renin levels (Woods, Pittman &

Pulliam, 1976; Stokes, Weber, Smith & Gain, 1976), and the inclusion of patients

taking diuretics (Bravo, Tarazi & Dustan, 1975).

Another explanation is the variability of the PRA measured on two different

occasions in the same patient. For instance, in 100 hypertensive hospitalized patients,

PRA was measured in the upright posture on the first and th ird days in hospital

(Figure 1), and the mean PRA values for each ofthese two days were respectively 1.39

± 1.06 and 1.47 ± 1.10 ng ml-1 h- 1

• The correlation between those two measurements

is even more significant (r = 0.793) than the correlation between the two measurements of 24h natriuresis (r = 0.413) . Although PRA values for patients can be

classified into three groups by comparison with control normotensive subjects, 28

out of the above 100 patients did not remain in the same group when PRA was

measured twice, even though the percentage of low, normal and high renin patients

remained identical.

These fluctuations around a 'cut-off' point obviously also apply to blood pressure

measurements. Individual sensitivity to ß -adrenoceptor blockade is far from being

the main component ofthe overall hypotensive response to ß-adrenoceptor blocking

drugs. These drugs have a moderate antihypertensive effect on most patients, and

clear demonstration of the differences observed from one individual to the othcr is

difficult. Variance analysis of blood pressure measurcments during a two day treatment of 50 hospitalized patients with acebutolol (patient factor, P) was performed.

Blood pressure was measured four times (reproducibility factor, R) before and four

r oO.793

1.39± 1.06 1.47 ± /

'"

1.10 PRA

(nq mi-I h-'j

UNoV

(rnrnol 24 h-1j

97±35 96 ± 32 "

/ r 00.413 r- -

-

35 56 91 33 58 91

Day I Day 3

Days 01 hospitalizafion

Figure 1 Renin subgroups in 100hypertensive patients: classification on the firstand third days

in hospital. Twenty eight patients transferred groups on the third day. The numbcr ofpatients in

each group is shown. The correlation between the two measurcmcnts of24h natriuresis is also

shown.

RENIN ANDß -ADRENOCEPTO R BLOCKING DRUGS 267

times after acebutolol absorption (treatme nt factor, T) . Systo lic blood pr essure in

recumbency decreased fro m 168 ± 17 to 14 5 ± 16 mmHg a nd diastoli c blood

pr essu re, fro m 106 ± 9 to 94 ± 10 mmHg. As sho wn in Table 2, on ly 13.2% of the

total va ria bi lity (sum of the squa re of the vari ance ana lysis) of systo lic pr essur e

measu rements and 16.6% of diasto lic measu rem ents dep en ded on ditTerences among

individua ls in the antihyperten sive etTects of ß-adren oceptor blocki ng drugs (patienttreatment interaction, PT).

Table 2 Varianee analysis of blood pressure measurements in 50 patients (P factor). Blood

pressure was measured four times (R factor) before and after treatment (T factor). Patienttreatment interaction is shown by a PT factor.

So urce of S ums 01 Degrees 01 Mean F

variation squa res freedom squares value

Diastolic blood P 25 509 49 521 13.0

pressure T 13 654 1 13 654 341***

PT 10391 49 212 5.3***

R 525 3 175

Error 12001 297 40

Systolic blood P 83229 49 1701 18.3***

pressure T 43326 I 43326 466***

PT 23 490 49 479 5.2***

R 400 3 133

Error 27 608 297 93

***p < 0.001

The anti-r enin effeet of ß-adre noceptor blocking dru gs: its par ticipation in the ir

antih yperte nsive property

Admi nistration of ß -adren oceptor blocking drugs to patients with mali gnant hypertension or stenosis of th e rena l artery reduced blood pressure with in a few hours

(Bühler et al.. 1975; Plouin , Me na rd, Corvol & Mi lliez, 1978 ), and this fall was

accom pa nied by a rapid decl ine in PR A. These c1inical ob servat ion s a re certai n ly th e

best evide nce indicating th at one wa y in which ß-adren oceptor blocking drugs exert

hypot en sive etTects is by inhibi ting the reni n-angiotensi n system. This conce pt is

strengthened by th e po sit ive correlation observed, in large gro ups of esse ntia l hyp erte nsive pat ients, bet ween the ci rcu lati ng level ofreni n a nd its decl ine under tre atment

o n th e one hand, and the fall in blood pre ssure on the other. Nev er the less, th e

co nce pt has two lim itations. Fi rstly, th e fall in ren in secretion ind uced by

ß -ad ren oceptor blocki ng dr ugs is far fro m accounting to co m plete elimi na tion ofthe

ren in- an giotensin system , neith er is there any evide nce ofintra rena l b lock ad e wit h its

conseq uent etTects on renal haem odynamics and tub ular sodi um reab sorption. This

is wh y other blocki ng age nts o f th e reni n-an giot en sin syste m will have to be fo und

(M en a rd & Corvol, 1980) . Secondly, th e ren in level m ight indicate so me disorder

wh ich do es not prima ril y in vo lve the ren in-an giotensin system. Thus, in a sma ll

gro up of selected hyperten sive pat ients, it was found th at th e fa ll in blood pressu re

induced by ß-adrenoceptor bloc king dru g tre atment corre lated just as weil with

plasma noradrenal ine measu red after 5 m in in the upright post ure as with PRA

(Plouin er al., 1979 ), whic h has also been obse rved in othe r stud ies, where plasm a

noradren al ine was measured during exercise (Di stler , Keim , Cordes, Phil ipp &

W ol tT, 1978 ). Sim ilarl y, in 10 severely hyper tensive patients, T hibonn ier er al. (in

prep aration) have ob served th at th e fall in blood pressure ind uced by the

ac ute blockade of co nverti ng-enzyme wit h ca ptopri l (SQ 14225, Squibb) (Img kg-I)

268 J. MENARD, P. BAUTIER, P. F. PLOUIN, M. THIBONNIER & P. CORVOL

correlated with both initial PRA and initial plasma vasopressin. The antivasopressin effect of this blockade of the renin-angiotensin system might therefore

partly account for its hypotensive action, just as weil as its anti-angiotensin effect.

These examples suggest to so me investigators that PRA could be only a 'rnarker' of

hypertensive patient responders to ß-adrenoceptor blocking drugs, without

necessarily involving the renin-angiotensin system in the pathogenesis ofthe increase

in blood pressure.

Conclusions

In seven successive trials, performed between 1974 and 1980, a highly significant

correlation between the fall in blood pressure induced by a cardioselective,

ß-adrenoceptor blocking drug, acebutolol, and the initial level of plasma renin

activity (PRA) was always observed. Renin measurement was performed in the

upright posture, after three days in hospital , under a diet containing 100 mmol

sodium daily.

The variability of PRA and blood pressure measured on different occasions in the

same patient explains why such correlations are useful to describe a phenomenon in

a population of low, normal and high renin patients, but have a limited value to

allow prediction of the efficacy of ß-adrenoceptor blocking drug treatment in the

individual patient.

References

Alhenc-Gelas , F., Plouin, P. F., Ducrocq, M. B., Corvol, P. & Menard, J. (1978). Comparison of

the antihypertensive and hormonal effects of a cardio-selective beta-blocker, acebutolol ,

and diuretics in essential hypertension . Am. 1. Med., 64, 1005-1012.

Birkenhäger. W. H., de Leeuw, P. W., Wester, A., Kho, T. L., Vandongen, R. & Falke, H. E.

(1977). Therapeutic elTects of beta-adrenoceptor blocking agents in hypertension. In

Advances in Inte rnal Medicine and Pediatrics, ed. von Herausgegeben, P., Frick, G. A.,

von Harnack, G. A., Martini , A., Prader, R., Shoen, R. & WollT, H. P. pp. 117-134.

Springer-Verlag: Berlin, Heidelberg.

Bravo, E. L., Tarazi, R. C; & Dustan , H. P. (1975). Beta-adrenergic blockade in diuretic-treated

patients with essential hypertension. New Eng. J. Med., 292,66-70.

Bühler, F. R., Burkart, F., Lutold, B. E., Kung, M., Marbet, G. & Pfisterer, M. (1975). Antihypertensive beta-blocking action as related to renin and age. A pharmacological tool to

identify pathogenetic mechanism in essential hypertension . Am. 1. Cardiol., 36,653-669.

Bühler, F. R., Laragh, J. H., Baer, L., Vaughan, E. D., Jr. & Brunner, H. R. (1972). Propranolol

inhibition of renin secretion: a specific approach to diagnosis and treatment of renindependent hypertensive diseases. New Eng 1. Med. . 287, 1209-1214.

Collste, P. & Haglund, K. (1979). Time-course of blood pressure decrease during initiation of

antihypertensive treatment with metoprolol. In 6th Scientifu: Meeting o{ the ISH, p, 126

Goteborg. June 11-13.

Distler, A., Keim, H. J., Cordes, U., Philipp, T. & Wolff, H. P. (1978). Sympathetic responsiveness and antihypertensive effect ofbeta-receptor blockade in essential hypertension . Am.1.

Med... 64,446-541.

Ducrocq, M. B., Degoulet, P., Charpentier, A., Corvol, P. & Menard, J. (1975). Le test a

l'acebutolol dans l'hypertension arterielle moderee. Action sur les intervalles de temps

systoliques. Nouv. Presse med.. 4, 3268-3272 .

Hansson, L. (1973). Beta-adrenergic blockade in essential hypertension. ElTects of propranolol

on hemodynamic parameters and plasma renin activity. Acta med. Scand., 550, 1-40 .

Kreft, c., Menard, J. & Corvol, P. (1979). Value of renin measurernent, saralasin test and

acebutolol treatment in hypertension . Kidnev Int .. 15, 176-183 .

Laragh, J. H. (1976). Modern system for treating high blood pressure based on renin profiling

and vasoconstriction-volume analysis: a primary role for beta-blocking drugs such as

propranolol. Am. J. Med., 61, 797-810.

Lehtonen, A. (1976). Atenolol in hypertension . Acta Ther.. 2,125-131 .

RENI N ANDß-ADRENOCEPTOR BLOCKI NG DRU GS 269

Meek ers, J., Missotten, A., Fagard, R., Dernuynck , D. , Harv engt , C, Pas, P., Billiet, L. &

Arner y, A. (1975). Pred icti ve value ofvariou s pa rameters for the antihypertensive elfect of

th e beta-blocke r IC166,082 . Arch. int. Pharmacodyn. Ther.. 213, 294-307.

Menard , J., Bertagna , X., N' G uyene, P. T ., Degoulet, P. & Corvol, P. (1976). Rapid

identification of patients with essential hypertension sensitive to acebu tolol. Am. J. Med..

60 ,886-890 .

Mena rd, J. & Corvol, P. (1 980): L'inhibition du systerne renine-angi otensine: un progres majeur

dans le trait em ent de l'Hypertension Arterielle? Nouv. Presse med..9, 579-583.

Plouin, P. F., Alhenc-Gelas, F., Degoulet, P., Corvol, P. & Menard, J. (1980). Importance and

limitation of"PRA measu rement for th e choice of initi al medication in 'stepped-ca re'

tre atment ofhyperten sion . Cardiovasc. Rev. Rep.. 1,31-3 7.

Plouin, P. F., Cornoy, E., Bohuon, c., Corvol, P. & Menard, J . (1 979). Activite renine et

noradrenaline plasmatiques a u cours du trait em ent antihypertenseur par l'acebutolol.

Nouv. Presse med..8, 1905-1909.

Plouin, P. F., Men ard , J., Corvol, P. & Milliez, P. (1978). Incidence des beta-bloquants sur

l'acti vite reeninc plasmatique. Point Actuel. Nouv. Presse med..7,2769-2774.

Stok es, G . S.. Weber, M. A., Smith, J. V. &Ga in,J. M. (1 976). Double-blind eross-over study of

propran olol and spironolactone in hypertension and plasm a renin measurements in

relation to therapeutic effects, In Systemic effects ofantihvpertensive agents, ed . Sarnbhi,

M. P., pp . 26 5-284. New York: Stratt on Intercontinental Medi cal Book Corpo ration.

Thibonnier, M., Lardoux, M. D. & Corv ol, P. (1 980). Compa rative trial of labetalol and

ace butolol, alone or associated with dih ydralazine , in treatment of essen tial hypertension.

Brit.J. clin. Pharmac.. 9, 561- 567.

Vaughan , E. D., Jr ., Laragh , J. H.. Ga vras, 1., Bühler, F. R., Gavras, H., Brunncr, H. R. & Baer ,

L. (1973). Volume factor in low and normal renin essential hyperten sion . Treatment with

either spironolactone or chlorthalidone. Am .1. Cardiol., 32 , 523-532.

Wo ods, J. W., Pittman, A. W. & Pulli am , C. C. (1976). Ren in pro filing in the treatment of

hypertension . Nclt'Eng 1. Atcd.. 294, 113 7-1143.

RATIONALAPPROACHES TO

CLINICAL THERAPY

A. ZANCHETTI

Istituto di Patologia Medica land Isti tuio di Ricerche Cardiovascolari,

Universitä di Milano, and Centro Ricerche Cardiovascolari. CNR , Milan, ltaly

Obviously, the most rational approach to therapy is to remove the cause of the

disease. However, not to speak of essential or idiopathic hypertension , th is ideal

approach is sometimes deceptive even in cases of secondary hypertension. This

explains why so many various and varying criteria have been formulated from time

to time for 'predicting' the outcome of renal artery surgery in renovascular hypertension , and why a patient with the correct pattern of renin activity in blood from

either renal vein, with a marked response ofboth blood pre ssure and plasma renin to

saralasin infusion or captopril administration, still can remain hypertensive after a

technically succe ssful correction of renal artery stenosis. Patients with Conn's

syndrome who remain or become hypertensive after ablation ofthe adenoma are not

infrequent, and hypertension has been reported to remain or reappear after abl ation

ofa solitary phaeochromocytoma.

In essential hypertension the situation is even more complicated. As there is no

cause to remove, the rational approach should consi st in matehing the pharmacological properties ofthe drug with the physiopathological alterations ofthe patient.

In doing that, two sets of difficulties are met, one relating to the patient, the other to

the drug.

As to the physiopathological profile of the patient this is certainly complex and

variable. Arterial pressure is regulated through a multiplicity of factors, and

measurement of even the main factors is hardl y conceivable. Inferring th e whole

physiopathological picture from a single variable or marker is unwarranted and

misleading. Some of these markers are of dubious significance, and indeed it is

unc ertain wheth er they really mean what we want them to mean.

Renin and catecholamines have been extensively discussed by Menard, Bautier,

Plouin, Thibonnier & Corvol (1980) and Birkenhäger & de Leeuw (1980), these

substances will be mentioned only to cast further doubts on the doubtful picture

provided in the previous papers.

The mea surement of renin was introduced for profiling the vasoconstriction

-volume spectrum of the hypertensive patient (Laragh, 1973), but no convincing

evidence has been brought forth to prove that volume is expanded in low renin

patients (Zanchetti, Stella, Leonetti, Morganti & Terzoli, 1976). Plasma renin

activity as currently measured after aperiod ofstanding upright, has also been interpreted as a sign ofsympathetic activity (Zan chetti, 1977). There is no doubt that the

increase in plasma renin in the upright posture is due to sympathetic stimulation of

renin secretion (Stella & Zanchetti, 1977; Zanchetti, 1977), but it is questionable

RAT10NAL APPROACHES TO CLINICAL THERAPY 271

whether the difTerences in renin level characterizing the subgroups of hypertensives

with high, normal and low renin are really indicating difTerences in sympathetic

activity. Indeed, Morganti, Pickering, Lopez-Ovejero & Laragh (1980) have shown

similar concentrations in plasma noradrenaline and adrenaline in the supine position

and similar increases during tilting in hypertensive patients classified as with high,

normal and low renin . This indicates that the difTerences in plasma renin in hypertensive patients may rather measure reactivity ofjuxtaglomerular cells than various

degrees ofsympathetic activation.

Plasma catecholamines have recently become increasingly popular as markers 01

sympathetic activity in hypertension. Birkenhäger & de Leeuw (1980) have just

summarized the many controversial sides ofthe problem.

It seems rational, before applying these indices in the rational approach to antihypertensive treatment, to ask ourselves whether plasma noradrenaline or adrenaline

are sufficiently sensitive indices of that mild increase in tonic sympathetic activity

that might be hypothetized to occur in essential hypertension or might difTerentiate

the various hypertensive patients among themselves (Zanchctti, Mancia & Leonetti,

1979). The well-known observation that plasma catecholamines increase during

tilting upright or exercise simply means that they can signal the tremendous

sympathetic activation occurring physiologically under these conditions. The point

is, however, whether plasma catecholamines can also reveal slight changes in

sympathetic activity.

Our fmdings cast some doubt on this possibility (Mancia, Leonetti, Picotti, Ferrari,

Galva, Gregorini, Parati, Pornidossi, Ravazzini, Sala & Zanchetti, 1979). In a group

of hypertensive patients tilting caused a two-fold increase in plasma noradrenaline,

and plasma adrenaline also rose. In the same subjects sympathetic activity was

modulated in both ways by stimulating or deactivating the carotid sinus baroreflex by

means ofthe neck pressure chamber (Ludbrook, Mancia, Ferrari & Zanchetti, 1977).

Reduction in baroreceptor activity caused a highly significant increase in mean

arterial pressure and in heart rate , but only a minor and nonsignificant increase in

plasma noradrenaline and adrenaline. Stimulation of the baroreceptors caused a

significant reduction in mean arterial pressure, but only a transient and nonsignificant decrease in plasma noradrenaline and adrenaline. This suggests that, if

difTerences in sympathetic activity among hypertensives are small, they may not be

adequately evaluated by using plasma catecholamines as markers of sympathetic

activity.

Finally, and obviously, the measurement of multiple variables with refmed

techniques, even more refined and rcliable than the ones available now, contrasts

with the requirement for simplified evaluation of the hypertensive patient,

especially the mild and moderate one, in view of the large numbers and the public

health problems involved.

If there are difficulties in correctly profiling the physiopathological pattern of the

individual patient, there are also uncertainties on the side of the drugs . 'No single

drug with a single action': this well-known axiom means that we are still confused

about the mechanism or the mechanisms ofaction ofseveral antihypertensive drugs .

Diuretics have been used in treating hypertension for more than twenty years, and

still their mechanism ofaction is disputed (Goldberg, Rick & Oparis, 1977): is it only

volume depletion or does volume gradually return toward normal and is a

vasodilator factor superimposed? The involvement of vascular wall prostacyclin in

the antihypertensive efTect of diuretics has just been suggested (Webster, Hensby,

Crowley & Dollery, 1980) and adds a further note to the key of possibilities to play

with .

p-adrenergic receptor blockers are also a splendid example of uncertain interpretation of certain efTects (Lewis, 1976). Their variable or dubious relations with renin,

catecholamines, central or peripheral sympathetic mechanisms, and with the haerno-

272 A. ZANCHETTI

dynamic pattern of hypertension have been discussed in previous papers, and need

not be reminded here in deta il.

Captopril is the latest example of our problems of interpretation . It has recently

been introduced as a specific agent for converting enzyme inhibition, and its excellent

antihypertensive action has been interpreted as entirely due to interference with the

renin-angiotensin system. It has indeed been found by most investigators that, on the

whole, the antihypertensive effect of captopril is related to plasma renin levels (Case,

Atlas , Laragh , Sealey, Sullivan & McKinstry, 1978), but evidence is accumulating

from different sources making the range of possible mechanisms wider than the

simple inhibition of converting enzyme. There is evidence, for instance, of time

dissociation between the hypotensive response and converting enzyme inhibition

(Waeber, Brunner, Brunner, Curtet, Turini & Gavras, 1980).

Ta mention only some recent observations of my group (Leonetti, Bianchini,

Terzoli & Zanchetti, in preparation), a single dose of captopril has been found

equally effective in lowering arterial pressure before and after one week administration of propranolol in doses (160 mg day-I) that had significantly reduced blood

pressure and largely suppressed plasma renin levels. It is interesting that

ß-adrenoceptor blockade, though suppressing base-line renin, did not interfere with

the rise in renin caused by captopril. Therefore, the preserved hypotensive response

to captopril after ß-adrenoceptor blockade cannot be ascribed to an interruption of a

renin feedback mechanism normally moderating the hypotensive response to

captopril.

These difficulties for a conceptual approach do not mean that a pragmatic

approach to treatment of hypertension should be an irrational one . Rationality can

be applied either for predicting the outcome of therapy before initiating it or for

correcting the kind or the size ofthe responses once occurred.

A good deal of improved predictability can certainly be derived from improved

knowledge of the physiopathology of hypertension obtained by more refmed

measurements in relatively small sampies of patients. Predictability, however, will

alwa ys be limited at the individual level. Even in ancient times, the Sibyl or the

Delphic oracle, used to speak by aphorisms, and it was at one 's own risk that the

aphorism was applied to one 's own future .

Rational correction ofinsufficient or unwanted responses once they have occurred

is a more rewarding approach, and can take undoubted advantage of the increasing

understanding ofthe physiopathology ofhypertension and the mechanisms ofaction

of antihypertensive drugs . For instance, we have shown that hypertensive patients

with a poor hypotensive response to diuretic treatment have an exaggerated increase

in renin as a consequence of the diuretic; the large blood pressure fall caused by

saral asin infusion in these patients indicates that it is this exaggerated increa se in

renin that blunts their hypotensive response to diuretics (Leonetti, Terzoli, Sala,

Bianchini, Sernesi & Zanchetti, 1978). Addition of a ß-adrenoceptor blocker to the

diuretic removes the counteraction of stimulated renin and produces a hypotensive

effect, with disappearance of the hypotensive response to saralasin previously

observed during administration ofthe diuretic alone. This supports the rationality of

adding a ß-adrenoceptor blocker to the diuretic in patients with poor response to

diuretics. Another well-known rational approach is that of adding in succession

ß-adrenoceptor blocker, diuretic and vasodilator in order to take advantage of the

mutual interferences between the various pharmacological properties ofthese drugs.

The great success achieved by antihypertensive therapy in the last twenty years has

almost inevitably given rise to some mythology, however. As with all myth s, those

about antihypertensive therapy also are distortions of well-founded concepts or

requirements.

One of these myths concerns the advantages of predicting the blood pressure

response to a given drug in a given patient. It is difficult to understand the real

RATIONAL APPROACH ES TO CLINICAL THERAPY 273

advantages of complicated and uncertain profiling for predicting the success of either

a diuretic or a ß-adrcnoceptor blocker in mild or moderate hypertension when the

slow pace ofthe disease gives all the time for trying one drug and eventually the other

one.

The other myth is an opposite one and is the public health myth of maximum

simplification, according to wh ich all hypertensive patients should be studied alike,

treated alike. This is not to deny the importance of guidelines for routine

examination and stepped-care prograrnmes, but to stress that all guidelines must be

intelligently and rationally applied to the individual patient. Likewise, the public

health problems created by the very large number of mild hypertensives should not

lead us to regard the more rare complicated patient as a nuisance for the major tasks

ofthe national health system.

It will not be inappropriate to conclude by mentioning that a rational approach to

therapy must necessarily be pragmatic. A rational pragmatic approach will take into

account what has been learnt from basic and clinical investigation about hypertension and antihypertensive drugs, as weil as what can be inferred from measurement of physiological variables in any given patient, but it will not try to substitute

prediction for monitoring and care , and will privilege careful readjustment of the

therapeutic regimen according to the favourable or unfavourable results obtained.

Finally, it will not forget that there are factors in the success of chronic therapy, such

as that of hypertension, that cannot be assessed or predicted by any metabolic or

haemodynamic measurement. There is no physiological concept, however attractive

and stimulating, than can make, for instance, a ß -adrenoceptor blocker weil

accepted by a heavy worker or clonidine by a truck driver.

The untoward etTects of drugs, so important in determining the patient's

compliance and the patient's quality of life, are not likely to be the same or to be

equally appreciated by every patient, by the old housekeeper and the retired

gentlemen as much as by the Latin Lover.

The judicious combination of physiological knowledge, clinical experience and

good horse-sense is likely to represent the safest guideline in any medical treatment,

antihypertensive therapy included.

References

Birkenhäger , A. H. & de Leeuw, P. W. (1980). Does the assay ofcatecholamine levels contribute

to the tailoring of antihypertensive treatment? In Proceedings 0/ the First World

Conference on Clinical Pharmacology & Therapeutics, ed. Turner, P., pp. 256-263.

London : Macmillan Publishers.

Case, D. 8., Atlas, S. A., Laragh, J. H., Sealey, J. E., Sullivan , P. A. & McKinstry, D. N. (1978).

Clinical experience with blockade ofthe renin-angiotensin-aldosterone system by an oral

converting-enzyme inhibitor (SQ 14.225, captopril) in hypertensive patients. Prog.

Cardiovasc. Dis.. 21, 195-206 .

Goldberg, L. 1., Rick, J. H. & Oparil , S. (1977). Pharmacology of antihypertensive agents: In

Hypertension. PhysiopathologyandTreatment. pp. 990-1024. New York: McGraw HilI.

Laragh, J. H. (1973). Vasoconstriction-vo1ume analysis for understanding and treating

hypertension: the use ofrenin and aldosterone profiles. Am. J. Med., 55,261-274.

Leonetti, G ., Terzoli, L., Sala, c.,Bianchini, c.,Sernesi, L.& Zanchetti, A. (1978). Relationship

between the hypotensive and renin-stimulating actions of diuretic therapy in hypertensive

patients. Clin. Sei. mol.Afed.. 55, 307S-309S.

Lewis, P. J. (1976). The essential action of propranolol in hypertension. Am. J. Med., 60,

837.

Ludbrook, J., Mancia, G., Ferrari , A. & Zanchetti, A. (1977). The variable pressure neck

chamber method for studying the carotid baroreflex in man . Clin. Sei. mol. Med., 53,

165-171.

274 A. ZANCHETT I

Mancia, G., Leonetti, G., Picotti, G. 8., Ferrari, A., Galva, M. D., Gregorini, L., Parat i, G.,

Pomidossi, G ., Ravazzin i, C; Sala, C. & Za nc hetti, A. (1 979). Plasma catecho lamines and

blood pressure respon se to th e carotid baroreceptor rellex in essential hypert en sion . Clin.

Sei.. 57 , 1655-167S.

Menard, J., Bautier , P., Plouin , P. F., Thibonnier , M . & Co rvol , P. (1980 ). ß -ad renoceptor

bloc king drugs and th e renin-an giot ensin system. In Proceedings 0/ the First World

Conference on Clinical Pharmacology & Therapeutics, ed. T urne r, P., pp. 264-269 .

Lon don : Macmillan Publishers.

Morganti, A., Pickering, T ., Lop ez-O vejero, J. & Laragh, J. H. (1 980). High and low ren in

subgroups of essential hypert en sion . Di fTerences and simi lar ities in their ren in and

sympathetic responses to neural and non -n eu ral stim uli. Am.1. Cardiol.. in press.

Stella, A. & Za nchett i, A. (1977 ). EfTects of renal den ervation on ren in release in respon se to

tilting and furosemide. Am. J. Physiol.. 232, H500-H507 .

Waeber, 8., Brunner , H. R., Brunner , D. B., Curtet, A. L., Turini, G . A. & Gavras, H. (1 980 ).

Discrepan cy between a ntihy pertensive efTect and angiotensin co nverti ng en zym e

inhibition by ca ptopril. Hypertension. 2, 236-242.

Webster . J., Hen sby, C. N. , Cr owley, K. & Doll ery, C. T. (1980). Circulating prosta cyclin levels

a re increased by bendrofluazide in patients with essential hypert en sion . 7th Sei. Me et. Im.

SOl'. Hypertension. New Orleans, p. 142.

Zan ch etti , A. (1977). Neural regulati on ofreni n release . Circulation, 56, 691--698.

Zan chetti, A., Man cia , G . & Leon ett i, G . (1979). Humoral markers of hyp ert ension. In

Radioimmunoassay of Drugs and Hormones in Cardiovascular Medicine. pp . 3-15.

Amsterdam : Elsevier /North Holl and.

Za nc hetti , A. , Stella, A ., Leonetti, G ., Morganti, A. & Terzol i, L. (1 976). Contro l of renin

release : a review of expe rime ntal evide nce and c1 in ical implica tio ns. Am. 1. Cardiol.. 37 ,

675--69 1.

Pharmacology and

Therapeutics of

Antirheumatic Drugs

Chairmen :

w. BUCHANAN, Canada

H. PAULUS, USA

THE THERAPEUTIC AND

TOXIC EFFECTS OF

ANTI-INFLAMMATORY DRUGS

WHICH INTERFERE WITH

ARACHIDONIC ACID

METABOLISM

G. A. HIGGS & B.J. R. WHITTLE

Department ofProstaglandin Research,

Wellcome Research Lab oratories,

Lan gley Court, Beckenham, Kent BR3 3B5, England.

Inflammation is a complex process wh ich presents several targets for drug action and

a large number of chemically unrelated drug s have bee n used in the treatment of

inflammatory diseases. There are, ho wever , signs and sym ptoms which are common

to inflammatory responses induced by different trauma. Furthermore, it is evident

th at man y anti-inflammatory drugs share th e same therapeutic and tox ic effects.

Van e (1 97 1) has suggested th at the anti-inflammatory effects and the gastro intestinal

irrita ncy of aspiri n-like drugs is du e to the inhibition of prostagiandin synthesis.

Evidenc e is now accumulat ing wh ich links the inhibition of arachidonic acid metaboli sm to the mech an ism of action of th e two rnajor groups of anti-inf1 a mmatory

drugs, the co rt icosteroids and th e non-steroid aspirin-like drugs. In th is paper these

theories will be reviewed and their implications for future drug development will be

discussed .

Metabolism of arachidonic acid

The discovery that polyunsaturated fatty acids such as arachidonic acid are precursors of prostaglandins has led to increased interest in lipid peroxidation. The

biochemical mechanism s ofprostaglandin synthesis have been elu cidated (for review

see Samuelsson , Goldyne , Gran strom, Hamberg, Hammarstrom & Malmsten, 1978)

and attention has now turned to other oxygenation products offatt y acid metabolism .

Arachidonic ac id is a substrate for cyclo-oxygenase and lipo xygenase enz ym es

(Figure I) (Hamberg & Samuelsson , 1974; Nugeren, 1975). Cyclo-oxygenase synthesises cyclic endoperoxides (PGG2, PGH2) wh ich are then transformed to stable

prostaglandins, thromboxan es or prostacyclin depending upon the tissue in wh ich

th ey are gene rated. These cyclo-oxygenase products have potent biological properties

and their pharmacology and endogenous roles ha ve been reviewed by Moncada &

Vane (1979).

278 G. A. HIGGS & B.J. R. WHITTLE

Lipoxygenases generate a number of open-chain mono- or di-hydroxy acids

(HETEs) from arachidonic acid via the corresponding unstable hydroperoxy acids

(HPETEs) (Figure I). 5-HPETE is also the precursor of the recently described

leukotrienes (Murphy, Hammarstrom & Samuelsson, 1979). The conjugation of

cysteine or a cysteine-containing peptide with the 5,6 epoxide of arachidonic acid

gives rise to a family of lipid-peptides which have potent biological activity. The

cyste inyl-glycinyl derivative of leukotriene A is identical to the slow-reacting

substance released in anaphylaxis (SRS-A) (Morris, Taylor, Piper & Tippins, 1980).

Arachidonic acid metabolism in inflammation

Cyclo-oxygenase

Prostaglandins are released whenever tissues are mechanically or chemically

stimulated and the y have been detected in numerous types ofinflammatory response

(for review see Ferreira & Vane, 1979). It is now c1earthat thromboxanes and

prostacyclin, as weil as the stable prostaglandins, are produced in inflammation. The

stable breakdown products of thromboxane A2 and prostacyclin (thromboxane B2

and 6-oxo-PGFIQ, respectively) have been found in comparable concentrations

with PGE2 in inflammatory exudates induced in rats (Higgs & Salmon, 1979).

Prostaglandins and thromboxanes have also been demonstrated in synovial fluids

aspirated from patients with rheumatoid arthritis (Trang, Granstrom & Lovgren,

1977).

The most likely source of inflammatory prostaglandins is the injured tissues

themselves but as the response proceeds, arachidonic acid metabolism could be augmented by the accumulation ofleucocytes. Cyclo-oxygenase has been found in every

mammalian tissue so far studied, with the exception of red blood cells . In blood

vessel walls the major cyclo-oxygenase product is prostacyclin (Moncada,

Gryglewski, Bunting & Vane, 1976) whereas granuloma tissues produce thrornboxanes and stable prostaglandins as weil as prostacyclin (Chang, Murota, Matsuo &

Tsurufuji, 1976; Chang, Murota & Tsurufuji, 1977). Arecent report shows that

human synoviaI tissue produces PGE2, thromboxane B2 and 6-oxo-PGFI Q

(Bitensky, Cashrnan, Chayen, Henderson, Higgs, Salmon & Vane, 1980). In this

study, PGE2 was the predominant product but thromboxane synthesis was higher in

tissues from rheumatoid patients than in tissues from patients with other inflarnmatory diseases.

Leucocytes are thought to be an important source ofthe cyclo-oxygenase products

found in inflammation. Phagocytosing polymorphonuclear leucocytes (PMNs)

release stable prostaglandins (Higgs, McCall & Youlten, 1975) and thromboxanes

(Higgs, Bunting, Moncada & Vane , 1976). Macrophages, which are the dominant

migratory cell in chronic inflammatory lesions also produce prostaglandins (Bray ,

Gordon & Morley, 1975), prostacyclin (Humes, Bonney, Pelus, Dahlgren , Sadowski,

Kuehl & Davies, 1977) and thromboxanes (Mu rota , Kawamura & Morita, 1978;

Brune, Glatt, Kalin & Peskar, 1978).

Lipoxygenase

Whereas the evidence for cyclo-oxygenase activity in inflammation is sub stantial,

there are few reports, as yet, of lipoxygenase products in inflammatory conditions.

There are, however, some indications that lipoxygenase act ivation has a role in

inflammatory responses both in experimental animals and man. 12-HETE (Figure I)

has been detected in the involved epidermis of patients with psoriasis and in this

condition the increase in lipox ygenase activity is far more evident than the increase

ANTI-INFLAMMATORY DRUGS 279

in cyclo-oxygenase activity (Harnrnarstrom, Harnberg, Sarnuelsson, Duell, Stawiski

& Voorhees, 1975). Furthermore, when arachidonic acid is injected into sites of

developing granuloma in rats, a significant proportion is converted to hydroxy acids

(Bragt & Bonta, 1979).

Siow reacting substances, which have now been shown to be lipox ygenase

products, are released in immediate hypersensitivity responses in man and animals

(for review see Austen, 1978). Leucocyte lipoxygenases produce a number ofhydroxy

acids as weil as the leukotrienes and stimulation of these cells with the calcium

ionophore A23187 selectively enhances peroxidation at the C-5 position of arachidonic acid (Borgeat & Samuelsson, 1979). Similarly, when sensitized human lung is

challenged with antigen there is an immediate release ofSRS-A (Orange , 1974). Thus

it is possible that lipoxygenase is activated in response to chemical or mechanical

irritation.

THROMBOXANE B2

TX B2

OO

6H

C17 HYDROXYACID

E 0 91 2 0P'

MALONDIALDEHYDE

MOA

OH

r\ ··~

HO"O OH

o

ciH 6H

PROSTAGlANDIN

OH

OOH o 'OH

PROSTAGlANDtN

6 OXO Fl",

6H 6H

OH

OOH OH OH

PROSTAGLANDIN F2.,.

/

LEUKOTRIENE C

(SRS) c: OH

C~ - I

-OH -

/

LEUKOTRIENE B

l5,12-DHETEI

c=i • c:!5::;

ARACHIDONIC ACID~ 5-HPETE 5-HETE

OO< crx:ki HOO HO

11 (12,15) HPETE 11 (12,15)HETE

CYCLIC ENDOPEROXIDES

r OOH 0'0·· < OoH

PGG2 !

PGI2 PROSTACYClIN .. OO THROMBOXANE A2

HOOC /" 0'0" < OH+O. <, OO PGH20 ~6H

OH 6H

METABOUSM CF ARACHIDONIC ACID

VIA CYClO-OX'lGENASE AND

UPOX'lGENASE ~THWAYS

Figure1 Metabolism of arachidonic acid via cyclo-oxygenase and lipoxygenase pathways.

280 G. A. HIGGS & B. J. R. WHITTLE

Intlammatory properties of cyclo-oxygenase products

The theory that prostaglandins are important inflammatory mediators has continued to gain support over the past ten years. Ryan & Majno (1977) have concluded

that prostaglandins should be grouped with vasoactive amines and kinins as the most

Iikely chemical mediators of vascular responses in acute inflammation. Prostaglandins of the E series and prostacyclin are potent vasodilators, producing a longlasting erythema following intradermal injection. This vasodilator property of cyclooxygenase products potentiates the oedema-producing capacity of mediators such as

histamine and bradykinin which increase vascular permeability (Williams & Peck,

1977). In contrast thromboxane A2 is a potent aggregator of platelets and a powerful

vasoconstrictor (Harnberg, Svensson & Samuelsson, 1975), properties which may

help to prevent haemorrhage in inflammation.

Systemic or local administration of stable prostaglandins is associated with the

development of hyperalgesia and overt pain (for review see Moncada, Ferreira &

Vane , 1978). Ferreira (1972) found that the pain-producing elTects ofbradykinin and

histamine are enhanced when alTerent pain-endings are sensitized by prostaglandins.

Prostacyclin is also hyperalgesic (Higgs, E. A., Moncada & Vane, 1978) and there is

evidence that prostacyclin rather than a stable prostagiandin accounts for the

hyperalgesia associated with carrageenin-induced inflammation in rats (Ferreira,

Nakamura & Abreu-Castro, 1978).

The role of prostaglandins in vascular responses in inflammation is well established but the elTect of cyclo-oxygenase products on inflammatory leucocyte function

is less well defmed. The stable prostaglandins influence leucocyte movement in so me

species but the accumulation of leucocytes at an inflammatory site is independent of

cyclo-oxygenase activity (Walker, Smith & Ford-Hutchinson, 1976). The 17-carbon

hydroxy acid HHT (Figure I) stimulates chemokinesis in T-lymphocytes whereas

other products of arachidonic acid ox ygenation are inactive (McCarty & Goetzl,

1979).

Elevation ofneutrophil cyclic AMP results in the inhibition ofmovement ofthese

cells and prostacyclin which is a potent stimulator of adenylate cyclase (Tateson,

Moncada & Vane, 1977) prevents leucocyte chemotaxis (Weksler, Knapp & Jaffe,

1977). Furthermore, prostacyclin inhibits the characteristic margination and

adherence of leucocytes in inflammed venules of the hamster cheek pouch (Higgs,

G. A. , Moncada & Vane , 1978). It is possible that prostacyclin modulates leucocyte

migration and it is interesting to note that total leucocyte numbers in inflammatory

exudates are at their highest when prostacyclin production is at its lowest (Higgs &

Salmon, 1979).