40

HEPAT 01050

Journal of Hepatology, 1992; 15: 40-47 @ 1992 Elsevier Science Publishers B.V. All rights reserved. 016%827g/92/$05.00

Efficacy of low-dose captopril in addition to furosemide and spironolactone in patients with decompensated liver isease during

blunted diuresis

A.A. van Vliet’, W.H. Hackeng3, A.J.M. Danker* and S.G.M. Meuwissen’

Departments of ‘Gastroenterology and 21nternal Medicine, Free University Hospital, Amsterdam and ‘IJsselland Ziekenhnis, Rotterdam, The Netherlands

(Received 3 December 1990)

The renjn-angiotensin-aldosterone system is activated by diuretics and involved in the diuretic resistance of cirrhotic

patients with ascites and oedema. In previous studies relatively high doses of captopril (25-400 mg daily) were unsuccessful in promoting diuresis and natriuresis in these patients. We analyzed the efficacy of a low dose of LdptOptil

in eight patients with massive ascites resistant to therapy of salt/fluid restriction and increasing doses of spironolactone and furosemide. Mean duration of diuretic use was 73 days (range 7-240 days). After at least 3 days of observation on 80 mg furosemide and 100 mg spironolactone only, captopril was added. Four out of eight patients responded with an increase in natriuresis and diuresis; daily dose of captopril was 20.6 mg in responders and 26.5 mg in non-responders. After the addition of captopril the mean weight change was -7.5 kg in responders and +0.25 kg in non-responders. Mean urinary sodium output in responders increased from 72.8 (S.D. = 35.2) to 128.5 (63.5) mmol within 10 days. Increased diuresis in responders made diuretic reduction necessary: mean furosemide from 80 to 53.3 mg, and mean spironolactone ffom 100 to 68.1 mg. Creatinine clearances remained stable. High levels of plasma renin activity, plasma aldosterone and angiotensin-II were found in all patients. Non-responders showed more severe hyponatremia and higher vasopressin levels. Natriuretic atria1 factor (NAF) was in the upper-normal range or slightly elevated in both groups. In non-responders we noticed low levels of cGMP in 24-h urine, compared with responders. Persisting high angiotensin-II and slightly elevated vasopressin levels may have impaired the function of NAF contributing to the persistently blunted diuresis in non-responders. Low-dose captopril provided a striking improvement of diuretic efficacy in 4/8 diuretic- resistant cirrhotic patients with ascites and oedema, and may, therefore, play an additional therapeutic role in these patients.

In patients with ascites and oedema due to decompen-

sated liver disease, salt restriction and treatment with diuretics are sometimes ineffective in removing excess of body fluid. Spironolactone and furosemide are the most widely used diuretics, but a randomized, controlled study in patients with ascites has never shown that this com- bination is the optimal regimen (1,2). Resistance to these agents, expressed by a stable weight with no increase in natriuresis and diuresis, makes treatment more compli- cated. The patient is in a so-called state of blunted di- uresis. Higher doses of diuretics are widely advised and

used with varying success, but include the danger of

generating encephalopathy or even hepatic coma (3). If high doses of diuretics fail to be effective, ascites punc- tions are done, but frequently the ascites is recollected after varying periods (4).

Early in the process of sodium and water retention the

renin-angiotensin-aldosterone system (RAAS) is in- volved (5,6). Activation of the RAAS is part of the

pathophysiology underlying cirrhosis and ascites (de- crease in effective arterial volume), but it is triggered by every attempt at volume depletion by diuretics, thus con-

Correspondence: Professor Dr. S.G.M. Meuwissen, Free University Hospital, Department of Gastroenterology, De Boelelaan 1117, 1081 HV Amsterdam, The Netherlands.

EFFICACY OF LOW-DOSE CAPTOPRIL 41

tributing to the development of diuretic resistance. In- tense activation of the RAAS makes diuretic require- ments greater. High plasma renin, angiotensin-II and aldosterone concentrations are found in such patients. The changed pattern of blood flow distribution in the cortex of the kidneys may contribute to the inefficacy of diuretics (7). Suppression of the RAAS by ACE- inhibitors appeared to be a logical step and has been tested by several investigators (8,9). Efficacy varied and dose was critical in maintaining good kidney function.

We have examined the influence of low-dose captopril in conjunction with fixed doses of furosemide and spironolactone to ascertain the exact influence of this angiotensin-I converting enzyme inhibitor on cirrhotic patients in a state of blunted diuresis. The captopril dose

was adjusted to blood pressure to minimize hypotension. This study was designed to answer the following ques-

tions: (i) Can low doses of captopril increase natriuresis and

diuresis in patients with cirrhosis of the liver when they

are in a state of blunted diuresis? (ii) Do low doses of captopril, when adjusted to blood pressure, have an impact on kidney function (as indi- cated by serum creatinine concentration)? (iii) Is there still a systemic hormonal effect of low doses

of captopril, as reflected by a decrease in the angio- tensin-II and aldosterone generation?

Patients and Methods

Eight patients (two female. six males) were studied.

All patients had massive ascites and seven patients had oedema of the legs. Ages varied from 40-79 years, with a mean of 57.7 years. Al patients had used spironolac- tone (~100 mg/day) uninterruptedly during 27 days. Five patients had been referred by other hospitals be-

cause of blunted diuresis during the use of furosemide 80 mg and spironolactone 100 mg per day (or more) for a

mean treatment period of 73 days (range 7-240 days). Exclusion criteria were: (i) creatinine clearance ~30

ml/min; (ii) systolic blood pressure <95 mmHg; (iii) cere- brovascular accident in the past; (iv) significant arrhyth-

mia or bilateral bundle branch block; (v) uncontrolled infection; (vi) active collagenic disease; (vii) leucopenia; (viii) obstructive cardiac valvular disease; and (ix) recent

myocardial infarction. Cirrhosis was of alcoholic origin in six out of eight

patients. One patient had cryptogenic cirrhosis and one

patient had developed cirrhosis after a hepatitis B infec- tion. According to the Child criteria, seven out of eight

patients belonged to group C.

Venous blood samples

Blood for measurement of plasma reniu activity,

angiOtenSin-11, plasma aldosterone, vasopressin and na- triuretic atria1 factor was drawn between &Of)-9:OO a.m.

after patients had fasted for 8 h and were recumbent for over 1 h. Blood samples were taken during the baseline period and every fourth day during concomitant use of captopril. Blood was collected in chilled tubes containing

sodium EDTA and was centrifuged immediately. Plasma was stored at -20°C. Urine was collected per 24 h for meaSUrement of electrolytes, creatinine, aldosterone and cyclic GMP.

Angiotensin-II (A-II) (N S-15 &ml) was extracted from plasma with ethanol. The supernatant was evapo- rated under a stream of nitrogen at 37°C. The hormone in the reconstituted extract was measured with a radio- immunoassay (RIA) kit of I.T.S. (Immuno Technology Service Production B.V., Wychen. The Netherlands).

Vasopressin (ADH) (N < 1.4 pg/ml) and natriuretic atrial factor (NAF) (11-68 pglml) were isolated from acidified plasma on Sep-Pak C-18 columns and eluated with acidified ethanol. The extracts were dried under a stream of nitrogen and the reconstituted residues were measured with the ADH RIB-kit (INCSTAR Sorin- Biomedica B.V. Amsterdam, The Netherlands) and the NAF RIA-kit of I.T.S.

Plasma renin activity (PRA) (N 0.3-3.5 ng A-I/ml) was measured as the production of angiotensin-1 (Ang-1) at 37°C during 1 h at pH 6. It was estimated in a RIA system with antiserum AS 130 (a gift from Dr. A.A.F. Hollemans, Academic Medical Centre, Amsterdam, The

Netherlands), human Ang-1 standard (SERVA Feinbio- chemica, Heidelberg, Federal Republic of Germany) and [1251]ANG-l tracer (NEN, Boston, MA, U.S.A.). The free and bound fractions were separated with a charcoal

dextran suspension. Plasma aldosterone (PA) (N 0.03-0.35 nmol/l) was

extracted with dichloromethane. The dried extract was measured after reconstitution in a RIA ll*ith [3H]aldo- sterone tracer and a charcoal-dextran suspension for sep-

aration of free and bound fraction. Urine aldosterone (N 11-55 nmoU24 h) was extracted

from acidified urine with dichloromethane. The dried

extract was redissolved in the lower layer of a Bush system water/methanol/petroleum~ther/toluene 3:7:5:5

and washed two times with the upper layer of this sys- tern.. Aldostersne was re-extracted from the lower layer with dichloromethane, dried with nitrogen and measured in a RIA, as mentioned for plasma aldosterone.

Cyc]ic-GMP (cGMP) (N 0.3-1.8 nmol/g Creatinine) in 24-h urine was measured without extraction in a dilution

of I:400 with a RIA (I.T.S.).

42

TABLE 1

Study medication, weight and blood pressure

Treatment Period

Furosemide

(mg)

Spironolactone Captopril

(mg) (mg)

Responders Diuretics Captopril

days 1-5 Captoprii

days 6-10 Captopril

days 11-15

Non-responders Diuretics Captopril

80.0 (0.0) 80.0 (0.0)

72.0 (16.4)

53.3 (25.7)

100.0 (0.0) 100.0 (0.0) 20.6 (10.2)

100.0 (0.0) 21.0 (10.8)

68.1 (31.9) 21.2 (11.4)

80.0 (0.0) 80.0 (0.0)

100.0 (0.0) 100.0 (0.0) 26.5 (8.3)

Weight

(kg)

64.0 (10.9) 62.0 (11.0)

58.2 (11.9)

104.0 (10.6) 67.7 (5.9) 79.8 (7.3) 105.8 (9.8) 65.8 (6.7) 79.1 (7.1)

102.0 (7.0) 65.5 (7.1) 77.7 (5.8)

56.5 (13.3) 97.7 (8.6) 60.0 Il0.0) 72.6 (9.0)

96.4 (13.9) 96.5 (16.2)

A.A. VAN VLIET et al.

Systolic blood pressure

(mmHg)

Diastolic blood pressure

(mmHg)

Mean arterial pressure (mmHg)

116.8 (8.2) 71.1 (7.4) 86.3 (5.9) 112.5 (10.3) 70.0 (9.7) 84.2 (9.4)

Mean (SD.).

Serum and urine electrolytes, and creatinine was mea-

sured by autoanalyzer methods.

Non-invasive recording

Ultrasonography was done to confirm the existence of ascites and to exclude impaired venous return in the inferior vena cava.

Statistical analysis

The analyses of this study are descriptive, given the sample size of eight patients. The patients were identified as responders (R; successful treatment with captopril) or non-responders (NR; unsuccessful treatment with capto- pril). These groups were compared on pertinent patient and disease characteristics.

The protocol had been approved by the Ethics Com- mittee of the Free University Hospital, Amsterdam, and all patients gave their informed consent.

Study design

All patients were hospitalized. After recruitment to our study all patients received fixed daily doses of 80 mg of furosemide and 100 mg of spironolactone, and ad- hered to a 50 mmol sodium (3 g NaCl per day) and 60 mmol potassium restricted diet. The patients had free access to water and bed rest was not obligatory. Basic

clinical variables were weight at 9:00 a.m., blood pres- sure t.i.d., fluid intake and 24-h urine volume. After an observation phase of at least 3 days on diuretics only, in addition to the treatment period elsewhere, captopril was

added when body weight had proven to be stable. Cap- topril was given at an initial dose of 3 mg at 6:00 a.m., noon and 6:00 p.m., and the dose was titrated to achieve

a haemodynamic response as determined by a fall in

systolic blood pressure of less than 15 mmHg. After the first dose, blood pressure was measured every 15 min during 2 h. On the other days, blood pressure was mea- sured three times daily and mean arterial pressure

(MAP) was calculated. Fluid balance was calculated daily from fluid intake and 24-h urine (from 8:00 a.m. to 8:00 a.m.). Net sodium output was calculated from sodium output minus sodium input minus 10 mmol (loss in the digestive tract).

Furosemide was given 40 mg orally b.i.d. and spirono- lactone 50 mg b.i.d. (at 6:OO a.m. and 6:00 p.m.).

Results

Eight patients completed the study. Four patients re- sponded, as evidenced by a remarkable increase in na- triuresis and diuresis on addition of captopril to the di- uretic regimen. An illustration of the evolving process is

given in Fig. 1. Four patients did not respond to capto- pril. The captopril dose ultimately achieved ranged from 3 to 12.5 mg t.i.d. with an average daily dose of 21 mg in responders and 26.5 mg in non-responders. Before admittance, the daily dose of furosemide ranged from

80-160 mg, while the daily spironolactone dose ranged from 100-200 mg. General body examination before the

study revealed mild encephalopathy in one in eight pa- tients.

Haemodynamics

Arterial pressure and mean arterial pressure are listed in Table 1.

EFFICACY OF LOW-DOSE CAPTOPRIL 43

Furosemide Sodium intake Vasopressin (ADH) ~O!nn10”z4 uur - N<1.4pgllM

I MO/

i ,4_- --____-___-! ! 12! n I

Body Weight

serum Sodium Natriuretic Atrial Factor

Blood Pressure

140~

cyclic GMP in 24 hrs urine

35, nrnolla lveet (0 II 8, 3 5

urine Sodium

Plasma Renin Activity b. N 0.353.5 ng AlIe0 “Ill

Aldosterone in 24 hours urine 600N I,-%mlK&ci

Angintensin II faON515F@ml

Aldosterone (blood) 6 ~03.0 35 “m&l

Fig. 1. Female. 40 years, chronic alcohol abuse, progressive oedema and ascites not reacting to sodium restriction. bed rest and uninterrupted therapy with spironolactone 100 mg and furosemide 40-80 mg daily during 14 days. At admission blood pressure was 95170 mmHg and weight 53 kg. After 6 days of observation on standard furosemide and spironolactone, captopril was added in a dose of 6.25 mg t.i.d. This was followed by a gradual development of natriuresis. Body weight decreased from 53 to 41 kg. Mean arterial pressure was hardly influenced.

Suppression of the RAAS was seen 2 h after captopril. but the 24-h aldosterone excretion remained 4-&fold of normal.

Ciinical parameters loss varied (l-6 days after the start of captopril).

Body weight decrease during the observation phase Increased diuresis and decrease in weight necessitated

was nil, but weight loss under treatment with captopril a reduction in diuretic requirements. The average daily

was remarkable. The responders lost all signs of oedema dose of furosemide was 80 mg before captopril and ul-

and ascites, and the maximum reduction in body weight timately 53.3 mg (range 80-20 mg) during concomitant

was 12 kg in 20 days (Fig. 2). Decreases in oedema and captopril therapy. The average daily dose of spironolac-

ascites were rapid, but the time of onset of the weight tone was 100 mg before captopril and decreased to 68.1

44

kg. 2-----------

-14------ --'F-------k-- -4 -2 0 2 4 6 6 la 12 14 16 18 20 22

day

Fig. 2. Effect of low-dose captopril addition to fixed doses of di- uretics in a phase of stable weight. With variance in the time of onset, body weight decreased and removal of ascites was reached.

mg (range 100-25 mg) after 10 days captopril addition

(Table 1). Mean fluid balance became clearly negative from day

1 to 5 of captopril use (Table 2). Mean (SD.) fluid intake was 1269 (327) ml daily before captopril. Fluid intake during captopril use was not much different, minimum intake during captopril coadministration being 1055 (312) ml on the days 11-15. Mean daily urine output was 1207 (456) ml before captopril and increased to a maximum of 1646 (395) ml on days 6-10 during concomitant use of captopril. None of the responders complained of exces- sive thirst during the dehydration. A transient feeling of fatigue was present, but all patients remained alert and kept attentive. An acute fall in blood pressure on the initial dose of 3 mg captopril never occurred. No patient had complaints related to hypotension. Captopril did not

have to be stopped in any patient, but a decrease in dose was necessary during titration on the blood pressure.

In the four patients who did not show a response to

TABLE 2

Balance data and clinical chemistry

Treatment period

Fluid Urine intake production (ml/24 h) (ml/24 h)

Responders Diuretics 1269 (3’7) Captopril days l-5 1100 (291) Captopril days 6-10 1156 (362) Captopril days 11-1s 1055 (312)

Non-responders Diuretics 1107 (334) Captopril 1375 (424)

1207 (456)

1413 (450)

1646 (395)

1262 (454)

941 (405) 1020 (392)

A.A. VAN VLIET et al.

the use of captopril, mean body weight rose 0.25 kg

(Table 1).

Laboratory variables After 3 days of baseline observation, serum electrolyte

levels and renai function remained essentially unchanged in all patients. Endogenous creatinine clearance (ECC) in all patients was greater than 3r ml/min, and ranged from 36-96 mumin at the start of the study. The mean net sodium excretion during the observation phase was ~13 mmoY24 h in all patients. On captopril t.i.d. this increased to 70.9 mmoV24 h in responders. Maximal

natriuresis was reached on t! e days 6-10, with a mean of 21 mg captopril in this period. Serum albumin levels ranged from 19 to 35 g/l with a mean of 23 g/l (Table 2).

During the study no hypokalaemia (serum K ~3.5 mmolfl) or severe hyperkalaemia (serum K >5.5) oc- curred. Before use of captopril, the average (S.D.) se- rum sodium was 130.7 (2.1) mmol/l in responders and during therapy with captopril serum sodium did not change. Values for potassium were 3.9 (0.8) and 4.3 mmol/l (0.7) respectively. Mean values of serum creat- inine changed from 91.7 (16.6) pmol/l in the pre-captopril period to 108.7 (26.9) PmoUl on days 6-10, when may- imal diuresis occurred.

Hormonal parameters (Table 3) Prior to administration of captopril, PRA was high in

all patients. Peak levels were present on days l-5 after captopril. A gradual decline in PRA was seen in the responder group. Mean A-II levels 2 h after administra- tion of captopril showed a progressive reduction during

the study, both in responders as well as in non- responders. PA 2 h after captopril showed the largest decrease on days 6-10 (from 2.5 to 0.9 pmoU1). The

Urine Urine Albumin Serum Serum Serum creatinine sodiuma (pn) sodium potassium creatinine (mmoU24 h) (mm01124 h) (mmoVI) (mmoUI) (qtol/l)

-_____

5.4 (2.6) 72.8 (35.2) 23.0 (5.4) 130.7 (2.1) 3.9 (0.8) 91.7 (16.6)

6.2 (2.4) 86.3 (46.7) 20.2 (7.0) 130.2 (3.8) 4.0 (0.4) 88.0 (29.3)

5.6 (2.7) 128.5 (63.5) 26.8 (6.3) 130.8 (2.1) 4.4 (0.6) 108.7 (26.9)

5.4 (1.8) 90.7 (46.2) 20.7 (4.2) 131.8 (2.3) 4.3 (0.7) 81.3 (30.8)

9.0 (1.7) 40.9 (35.0) 24.4 (3.7) 128.5 (6.3) 4.5 (1.0) 116.1 (2.9) 9.8 (2.3) 42.4 (32.0) 26.0 (4.9) 126.7 (5.5) 4.5 (0.6) 132.7 (31.2)

Mean (S.D.). “Sodium intake 50 mmollday.

EFFICACY OF LOW-DOSE CAPTOPRIL

iABLk 3

Bormone data

Treatment Plasma renin period activity

(ng Angio-I/ml)

-___ Responders

Diuretics 26.0 (20.0) Captopril 43.0 (37.0)

days 1-5 Captopril 32.9 (18.1)

days 6-10 Captopril 28.2 (16.7)

days 11-15

Non-responders Diuretics 22.8 (16.1) Captopril 42.8 (14.8)

Mean (S.D.).

Angiotensin-II Aldosterone (pg/ml) in blood

(nmolll)

194.0 (140.8) 2.5 (2.0) 147.0 (176.6) 1.7 (1.0)

96.5 (99.8) 0.9 (0.3)

68.5 (63.0) 1.3 (0.7)

174.7 (104.7) 1.9 (1.0) 117.0 (56.2) 1.4 (0.4)

reduction in mean urinary aldosterone was 25% on days

6-10, and rose to baseline values after complete removal of ascites. Plasma ADH was within the normal range and did not change during the study. Basal levels in non- responders, however, were 1.3 (0.4) pg/ml vs. 0.7 (0.1) in responders. Mean levels of NAF were elevated in responders and non-responders. During dehydration a gradual fall in NAF was noted. Mean urinary cGMP in responders was above normal and corresponded with

NAF, but urinary cGMP did not parallel the decrease of the NAF in blood during dehydration. In the non-

responders we found a low-normal level of cGMP in the presence of elevated NAF values. During captopril, mean NAF in non-responders rose, while cGMP in the

urine remained essentially unchanged.

Adverse effects

During the study captopril was well tolerated. No se-

rious neurological sequelae occurred. No patient devel- oped a marked decrease in MAP, renal insufficiency,

rash or blurred vision.

Follow-up The follow-up of the patients who were successfully

treated showed, that tapering of the ACE inhibitor im- mediately led to retention of fluid and an increase in

weight in two patients. One of these patients complained of a dry cough and generalized itching over the body

after 10 months at 18.75 mg captopril daily. Blood chem- istry was completely normal. The other patient was sta- ble on diuretics plus captoprii, but blood pressure tended to decrease to 70150 mmHg on continuous use of capto- pril 6.25 mg t.i.d. For this reason the dose was changed

after 6 months to 6.25 mg twice a day. Two patients were

45

Aldesterone in ADH NAF cGMP urine (pg/ml) (pdml) in 24-h urine (nmol/2i h) (nmol/g

creatininc)

123.7 (154.1) 0.7 (0.1) 83.7 (53.2) 114.3 (101.0)

1.93 (1.43) 0.7 (0.3) 72.8 (31.6) 1.80 (1.95)

92.1 (55.7) 0.8 (0.2) 49.2 (20.5) 1.72 (1.02)

136.2 (96.6) 0.8 (0.1) 49.5 (15.6) 1.88 (0.92)

409.5 (338.4) 1.3 (0.4) 83.1 (40.5) 0.82 (0.26) 171.9 (77.0) 1.1 (0.2) 106.5 (69.3) 0.84 (0.26)

weaned off captopril treatment without problems. In one

patient a low serum sodium concentration (cl28 mmol/l) during captopril persisted, but fluid restriction corrected this to 132 mmohl. The other three patients had levels >132 mmol/l on follow-up.

bcession

This study in a limited number of cirrhotic patients with persistent ascites and oedema shows, that addition of a low-dose of captopril to a fixed combination of furosemide and spironolactone reversed blunted diuresis in 50% of the patients without serious side effects. The

beneficial effect was probably due to the effects on the BAAS, which was highly activated either on pathophys- iological grounds or by previous diuretic use.

One a slight decrease in glomerular filtration during captopril administration was observed in periods of gross natriuresis and diuresis in all responders. Some non- responders also showed a slight decrease in glomerular filtration during stable natriuresis. Serum electrolytes were not greatly influenced by captopril coadministration and severe hyperkalaemia never occurred. We demon-

strated that even 3 mg of captopril can induce a hor- monal shift. Low-dose captopril led to submaximal A-II suppression and exerted a transient effect on the PA. Overall reduction in the 24-h aldosterone excretion was 25% compared to baseline values (use of diuretics only). Thus, an explanation of the increased natriuresis during captopril administration based strictly on aldosterone suppression is unlikely. In the literature conflicting data exist on the role of captopril in the treatment of ascites in liver cirrhosis. Captopril itself has no strong diuretic

46 A.A. VAN VLIET et al.

action. The natriuretic effects which we observed, should be explained by improved efficacy of the prescribed di- uretic combination. The phenomenon was documented earlier by Sheperd et al. and Jorgensen et al. (10,ll): high doses of diuretics (160 mg of furosemide and 400 mg of spironolactone) in patients with refractory ascites were made effective by addition of 6.25 mg of captopril t.i.d.

These patients, however, showed transient but severe

signs of drowsiness until captopril was withdrawn. In our

patients, the mental state of the patients was not influ- enced by captopril. Others have reported a decrease in GFR, a decrease in natriuresis, and development of hy- potension as side effects of captopril administered as a single dose of 25-150 mg to patients with liver cirrhosis

(89). Wood et al. reported a significant fall in MAP associated with side effects such as postural hypotension, reduced GFR and increasing encephalopathy during cap- topril therapy at a dose of 6.25 on 12.5 mg every 8 h (12). The main differences between those studies and our study are the dose and the fact that our patients were studied for a period longer than a single day. Our pa- tients could thus compensate for the initial effects of captopril. Recently, Brunkhorst and co-workers (13) published the effect of captopril (6.25 mg four times daily) in cirrhotic patients with ascites. An increase in natriuresis in 1203 patients with diuretic resistance was observed. In that study, doses of diuretics varying from very low to very high were used. Side effects of captopril administration were absent.

Several mechanisms might be responsible for the cap- topril effect. In cirrhotic patients with ascites there is vasodilation in the splanchnic and extrasplanchnic areas which corresponds to activation of the RAAS and re- duction in renal blood flow. Activation of the RAAS can

result in a shift from preserved renal function to func- tional renal failure and azotaemia. Both high systemic A-II level and local production of A-II within the kidney

might be responsible for renal vasoconstriction (14-18). Many investigators have suggested that local intrarenal effects of captopril are important. Captopril can redirect the blood flow in the cortical areas of the kidneys by creating vasodilation (7,8). In addition, renal vasodila- tors like prostaglandins and kinins, which are low in cirrhotics (19,20), especially in those with functional re-

nal failure, might be increased by captopril (21-23). This local vasodilatory action in the kidneys after captopril

administration might contribute to the improved diuretic

response. From a pharmacological point of view, the addition of captopril to our diuretic regimen is similar to

a method known as sequential nephron blockade (24). In this method an agent with carboanhydrase inhibitory ac- tivity interferes with the most proximal tubular reabsorp-

tion of sodium and shows synergism with the combina- tion of a loop diuretic and a potassium-sparing agent like spironolactone, which both prevent sodium reabsorption at more distal sites of the tubule. Stimulation of the RAAS is known to increase tubular reabsorption of so- dium at the most proximal tubular parts of the nephron under influence of a high level of A-II. During blunted diuresis in our patients, relatively low doses of furo- semide and spironolactone had no effect. This was prob- ably due to increased proximal sodium reabsorption. As a result the load to the distal tubule was low. Captopril administration reduced the A-II effect in the proximal tubule, increased the soluble load to the distal tubular parts, and enabled furosemide and spironolactone to pre-

vent sodium reabsorption, thus increasing natriuresis. Captopril also interferes in the balance between NAF

and A-II. A high level of A-II in the kidney from sys- temic origins or locally generated, could be a cause of the

inefficacy of NAF (25). In our study we found some support for this hypothesis. The cellular effects of NAF are thought to be mediated, at least in part, by cyclic 3’,5’-guanosine monophosphate (cGMP) as a second messenger, by activation of the enzyme guanylate cy- clase. A linear correlation exists between NAF concen- trations in the blood and cGMP concentrations in urine (26-28). Therefore, cGMP is regarded as a biologic marker for the cellular response to NAE In the patients responding to captopril we found increased NAF con- centrations in the blood and high-normal urinary cGMP

levels during baseline observation. A gradual decrease in NAF occurred when diuresis developed under captopril. The urinary cGMP in the responders remained high dur-

ing augmented natriuresis. Improved cellular response to N AF might explain this phenomenon. In non-responders we found increased levels of NAF, but cGMP concen-

trations in urine were remarkably low. During captopril administration cGMP remained low with a low natriure- sis. Insufficient biologic expression of NAF by persisting

high A-II might be one of the factors which can explain decreased natriuresis in non-responders.

Although doubt exists concerning long-term treatment results with low doses of captopril in liver disease, cer- tain patients did not experience any side effects. In those

cases diuretic efficacy and salt tolerance (and thus dietary freedom) improved remarkably by co-administration of captopril .

Our study tested the natriuretic effects of captopril when administered with relatively low doses of diuretics.

In refractory ascites, however, doses of up to 160 mg of furosemide and 400 mg of spironolactone are widely

used. If these larger doses were made effective by the addition of captopril the mobilization of oedema and

EFFICACY OF LOW-DOSE CAPTOPRIL 47

ascites might be stronger and faster than in our respond- ers for whom diuretics were tapered due to progressive diuresis. Theoretically it might be better to add low doses

of captopril to (relatively) low doses of diuretics and to increase the diuretic dose if needed, than to add capto- pril to maximal doses of diuretics during blunted diure- sis. This avoids the chance of sudden, massive diuresis with side effects like life threatening encephalopathy or renal failure, if effective arterial volume is decreased too drastically.

In conclusion, our study shows that the addition of low dose captopril to relatively low doses of diuretics has the potential to reverse blunted diuresis in cirrhotic patients

without ivlduction of hepatic encephalopathy, renal im- pairment or electrolyte disturbances. The traditional ap- proach in patients with cirrhosis and insufficient diuretic response has been an increase in the doses of diuretics,

References

1 Arroyo V, Rodes J. A rational approach to the treatment of ascites. Postgrad Med J 1975; 51: 558-62.

2 Descos L, Gaulthier A, Levy VG, et al. Comparison of six treatments of ascites in patients with liver cirrhosis. a clinical trial. Hepato-gastroenterol 1983; 30: 15-20.

3 Sherlock S, Senewiratne B, Scott A, Walker JG. Complications of diuretic therapy in hepatic cirrhosis. Lancet 1966; i: 1049-53.

4 Gines P, Tito L, Arroyo V, et al. Randomized comparative study of therapeutic paracentesis with and without intravenous albumin in cirrhosis. Gastroenterology 1983; 94: 1493-502.

5 Arroyo V, Bosch J. Mauri M, et al. Renin, aldosterone and renal haemodynamics in cirrhosis with ascites. Eur J Clin Invest 1979; 9: 69-73.

6 Wilkinson SP, Williams R. Renin-angiotensin-aldosterone sys- tem in cirrhosis. Gut 1980; 21: 545-54.

7 Merkel C, Gatta A, Milani L, Amodio P. Zuin R. Intrarenal blood flow, circulation time and cortical vascular volume in patients with cirrhosis. Stand J Gastroenterol 1981: 16: 775-80.

8 Pariente EA, Bareille E, Bercoff D, Lebrec D. Acute effects of captopril on systemic and renal hemodynamics in cirrhotic pa- tients with ascites. Gastroenterology 1985; 88: 1255-9.

9 Daskalopoulos G, Pinzani M, Murray N, Hirschberg R, Zipser RD. Effects of captopril on renal function in patients with cir- rhosis and ascites. J Hepatol 1987; 4: 330-6.

10 Sheperd AN, Neligan P. Hayes PC. Captopril and resistant ascites. Lancet 1983; i: 1391-2.

11 Jorgensen F, Badskjaer J, Nordin H. Captopril and resistant ascites. Lancet 1983: ii: 405.

12 Wood LJ, Goergen S, Stockigt JR, Powell LW, Dudley FJ. Adverse effects of captopril in treatment of resistant ascites, a state of functional bilateral renal artery stenosis. Lancet 1985; ii: 1008-9.

13 Brunkhorst R, Wrenger E, Kuehn K. Schmidt RN, Koch K. Effekte einer Captopriltherapie auf die Natrium- und Wasserausscheidung bei Patienten mit Leberzirrhose und Aszites. Klin Wochenschr 1989; 67: 774-83.

14 Lohmeier TE, Cowley AW Jr, Trippodo NC, Hall JE, Guyton AC. Effects of endogenous angiotensin II on renal sodium ex- cretion and renal hemodynamics. Am J Physiol 1977; 233: 388- 95.

15 Jin M, Wilhelm MJ, Lang RE, Unger T, Lindpaintner K, Ganten D. Endogenous tissue renin-angiotensin systems. Am J

thus possibly stimulating the RAAS to formidable heights and further inducing diuretic resistance. Investigation is

needed to show if application of a low dose ACE inhib- itor in early stage ascites can be used to prevent diuretic resistance in later stages.

Dr. W.W. Meyer (St. Geminiziekenhuis), Dr. G.H. Krull (Burgerziekenhuis, Amsterdam) and Dr. J.9.M.

Vogten (Spaarneziekenhuis, Heemstede) are kindly ac- knowledged for patient referral. The results of this study were presented as Poster at the World Congress of Gas- troenterology, August 1990, Sydney. The study was sup- ported by grants of SQUIBB and CEDOPHAR B.V.

Med 1988; 84 (Suppl. 3A): 28-36. 16 Baillie MD, Rector FC Jr, Seldin DW. Angiotensin II in arterial

and venous plasma and renal lymph in the dog. J Clin Invest 1971; 50: 119-26.

17 Baillie MD, Loutzenhieser R, Moyer S. Relation of renal he- modynamics to angiotensin II in renal hilar lymph of the dog. Am J Physiol 1972: 222: 1075-8.

18 Disalvo J, Peterson A, Montefusco C, Menta M. Intrarenal conversion of angiotensin I to angiotensin II in the dog. Circ Res 1971; 29: 398-406.

19 Perez-Ayuso RM, Arroyo V. et al. Renal kallikrein excretion in cirrhotics with ascites: relationship to renal hemodynamics. He- patology 1984; 4: 247-52.

20 Rector WG. Urinary prostaglandin E2 excretion, sodium reten- tion and diuretic responsiveness in patients with chronic liver disease. Am J Gastroenterol 1987: 82: 347-51.

21 Swartz SL, Williams GH, Hollenberg NK. Captopril-induced changes in prostaglandin production: relationship to vascular responses in normal man. J Clin Invest 1980; 65: 1257-64.

22 Zusman RM. Effects of converting-enzyme inhibitors on the renin-angiotensin-aldosterone. bradykinin, and arachidonic acid- prostaglandin systems: correlation of chemical structure and bi- ologic activity. Am J Kidney Dis 1987; X (Suppl.): 13-23.

23 Dzau VJ, Swartz SL. Dissociation of the prostaglandin and renin angiotensin systems during captopril therapy for chronic congestive heart failure secondary to coronary artery disease. Am J Cardiol 1987; 60: 1101-5.

24 Puschett JB. Clinical pharmacologic implications in diuretic se- lection. Am J Cardiol 1986: 57: 6A-13A.

25 Showalter CJ, Zimmerman RS, Schwab TR, Edwards BS. Op- genorth TJ. Burnett JC Jr. Renal response to atria1 natriuretic factor is modulated by intrarenal angiotensin II. Am J Physiol 1988; 254: 453-6.

26 Lewis H, Wilkins M, Selwyn B, Yelland U, Griffith M. Bhoola KD. Relationship between ANP, cyclic GMP and tissue kal- likrein following saline infusion in healthy volunteers. Adv Exp Med Biol 1989; 247A: 281-6.

27 Hamet P, Tremblay J, Pang SC, et al. Effect of native and synthetic atria1 natriuretic factor on cyclic GMP. Biochem Bio- phys Res Commun 1984; 123: 515-27.

28 Gerzer R. Witzgall H, Tremblay J, et al. Rapid increase in plasma and urinary cyclic GMP after bolus injection of atria1 natriuretic factor in man. J Clin Endocrinol Metab 1985; 61: 1217-9.