Interaction of spironolactone with ACE inhibitors or angiotensin receptor blockers: analysis of 44 cases

The randomised aldactone evaluation study (RALES) proved a substantial (30%) reduction in risk of mortality in patients with severe congestive heart failure by treatment with low dose spironolactone (25-50 mg a day) in addition to standard treatment.¹ Exclusion criteria for treatment in the study were a plasma potassium concentration > 5.0 mmol/l and serum creatinine concentration > 221 μmol/l. A pilot study had previously shown that the higher the dosage of spironolactone (up to 24% with 75 mg a day) the higher the risk of hyperkalaemia.² Standard treatment for patients with heart failure categorised as New York Heart Association class II to IV includes angiotensin converting enzyme (ACE) inhibitors or angiotensin II AT₁ receptor antagonists (AT₁ receptor blockers).³ Both spironolactone and ACE inhibitors or AT₁ receptor blockers reduce the renal elimination of potassium.⁴ In RALES, the increase in potassium was judged not to be important as serious hyperkalaemia (> 6 mmol/l) occurred in only 10 (1%) of 841 patients taking placebo and in 14 (2%) of 822 patients taking spironolactone, with no significant difference between the groups. Discontinuation of the treatment was necessary in only one patient taking placebo and three patients taking spironolactone.¹

We present a larger case series of life threatening hyperkalaemia in patients who were receiving spironolactone plus ACE inhibitors or AT₁ receptor blockers. We identify clinical circumstances associated with this medical emergency and suggest recommendations for prevention.

Case series

From January 1999 until December 2002 we observed 44 patients (17 men) with congestive heart failure who were taking spironolactone and ACE inhibitors or AT₁ receptor blockers and were admitted to our nephrology unit (serving a population of about 250 000) for treatment of life threatening hyperkalaemia. Their mean age was 76 (standard deviation 11) years. The mean dosage of spironolactone was 88 (SD 45, range 25-200) mg daily. All patients also received ACE inhibitors or AT₁ receptor blockers (table). Fourteen patients were treated with β receptor blockers and 40 with loop diuretics.

Table 1.

Clinical data for 44 patients with heart failure treated with combination of spironolactone plus ACE inhibitors or AT₁ receptor blockers

Patient No	Age (years)	Type 2 diabetes	NYHA class	Spironolactone (mg/d)	ACE inhibitor or AT1 receptor blocker (mg/d)	Potassium (mmol/l)*	Serum creatinine (μmol/l)*	Creatinine clearance (ml/s)*†	Initial treatment	Potassium (mmol/l)‡	Creatinine (μmol/l)‡	Creatinine clearance (ml/s)‡	Outcome
1	87	Yes	III	100	Enalapril (10)	7.09	165	0.42	C	3.40	122	0.57
2	88	No	III	50	Captopril (37.5)	8.50	227	0.27	C	5.00	79	0.77
3	86	Yes	IV	100	Losartan (50)	8.50	161	0.55	HD	4.80			Long term HD
4	88	Yes	IV	200	Ramipril (1.25)	8.27	363	0.22	HD	4.50			Long term HD
5	69	Yes	III	100	Ramipril (5)	7.80	201	0.50	HD	4.80	165	0.62
6	74	Yes	III	100	Benazepril (20)	9.10	138	0.63	HD	3.60	133	0.65
8	79	Yes	III	100	Benazepril (5)	7.20	171	0.60	HD	5.40	152	0.67
8	66	Yes	IV	100	Enalapril (5)	6.40	394	0.22	HD	4.50			Long term HD
9	67	No	IV	50	Enalapril (5)	8.04	447	0.25	HD	3.94			Long term HD
10	66	Yes	III	50	Losartan (50)	7.96	108	1.12	HD	3.73	80	1.50
11	66	Yes	III	50	Losartan (50)	6.20	215	0.50	HD	4.02	133	0.80
12	69	No	III	50	Captopril (50)	8.00	750	0.13	HD	4.75	125	0.78
13	73	Yes	III	50	Losartan (50)	7.50	126	0.60	HD	5.13	98	0.78
14	56	No	IV	50	Captopril (50)	7.50	180	0.73	HD	4.20	145	0.90
15	90	Yes	III	50	Enalapril (10)	6.30	109	0.50	C	4.49	120	0.45
16	50	Yes	IV	50	Enalapril (10)	7.50	594	0.27	HD	3.80	153	1.02
17	78	Yes	III	50	Benazepril (5)	7.40	126	0.65	HD	5.20	88	0.92
18	77	Yes	III	50	Moexipril (10)	8.40	185	0.42	HD	4.35	195	0.40
19	64	Yes	III	150	Enalapril (10)	6.73	231	0.37	HD	4.22	103	0.83
20	88	Yes	III	50	Captopril (50)	6.80	192	0.35	HD	4.47			Death
21	83	No	II	100	Captopril (50)	7.36	462	0.18	HD	4.49	121	0.72
22	75	Yes	II	100	Enalapril (5)	7.60	478	0.23	HD	4.30			Death
23	51	Yes	III	50	Enalapril (5)	7.32	295	0.59	HD	4.60			Long term HD
24	89	Yes	III	100	Captopril (12.5)	6.04	304	0.18	C	4.50	220	0.27
25	76	Yes	III	100	Captopril (150), telmisartan (80)	8.66	358	0.23	HD	3.73	92	0.90
26	81	Yes	III	200	Ramipril (5)	6.67	88	1.00	HD	4.84	109	0.80
27	76	Yes	II	50	Enalapril (10)	7.40	548	0.17	HD	3.51	242	0.38
28	67	No	III	25	Enalapril (20)	8.05	517	0.20	HD	4.03	234	0.45
29	68	Yes	III	100	Enalapril (10)	8.19	288	0.43	HD	4.83	106	1.18
30	70	Yes	III	50	Benazepril (20)	6.20	165	0.68	C	3.71	119	0.95
31	85	Yes	III	150	Captopril (50)	6.50	242	0.27	C	4.18	117	0.55
32	76	Yes	III	100	Captopril (50)	7.20	231	0.35	HD	4.22	105	0.78
33	88	Yes	III	50	Enalapril (10)	7.80	288	0.22	HD	3.89	155	0.40
34	91	Yes	III	100	Captopril (50)	7.10	380	0.20	HD	4.65	180	0.42
35	69	No	II	100	Captopril (50)	7.60	340	0.28	HD	4.44	103	0.92
36	73	Yes	III	50	Enalapril (10)	6.80	195	0.59	HD	4.50	110	1.05
37	81	No	IV	50	Enalapril (10)	7.80	266	0.32	HD	4.29	241	0.35
38	73	Yes	III	100	Captopril (75)	9.36	144	0.50	HD	4.56	97	0.75
39	63	Yes	II	150	Candesartan (16)	9.65	940	0.13	HD	4.64	306	0.42
40	90	Yes	III	150	Enalapril (5)	7.19	156	0.32	C	4.99	89	0.57
41	90	Yes	III	100	Enalapril (5)	8.74	299	0.17	HD	5.30	131	0.38
42	73	Yes	III	100	Enalapril (400	8.94	288	0.25	HD	3.45			Long term HD
43	83	Yes	III	50	Enalapril (10)	8.97	246	0.27	HD	3.74	168	0.38
44	76	No	III	200	Captopril (150)	8.75	319	0.27	HD	4.73	246	0.33

NYHA=New York Heart Association; HD-haemodialysis; C=conventional potassium lowering treatment (see text).

^*On admission.

^†Calculated by the Cockroft and Gault formula (for SI units).

^‡At last follow up.

Thirty five of the 44 patients had diabetes mellitus type 2. Symptoms on admission were vomiting (19), diarrhoea (8), bradyarrhythmia (14), muscle weakness and paralysis (27), and severe dehydration (28). Mean plasma potassium concentration on admission was 7.7 (SD 0.9, range 6.04-9.65) mmol/l and mean serum creatinine concentration was 294 (SD 175, range 88-940 μmol/l. Nineteen of the 44 patients had a serum creatinine concentration < 221 μmol/l. Creatinine clearance estimated by the formula of Cockroft and Gault⁵ was 0.38 (SD 0.22) ml/s and below the lower limit of normal in all patients. Five patients had to be resuscitated on admission. The figure shows electrocardiograms for patient 39, who on admission had a serum potassium concentration of 9.65 mmol/l and, after dialysis, had a lowered value of 4.65 mmol/l.

Figure 1.

Two electrocardiograms (ECGs) of a 63 year old patient (patient 39 in table 1). The top ECG (A) was recorded when the patient presented with muscle paralysis, bradyarrhythmia (25-35 beats/min), hypotension (blood pressure 60/30 mm Hg), and a serum potassium concentration of 9.65 mmol/l. The patient improved rapidly after immediate treatment with haemodialysis in the intensive care unit. The bottom ECG (B) was recorded six hours later, when the patient's serum potassium concentration was 4.65 mmol/l.

Haemodialysis was immediately started in 37 patients. Seven patients received conventional potassium lowering treatment consisting of intravenous sodium bicarbonate, intravenous furosemide (frusemide), and intravenous glucose and insulin. The mean number of dialysis treatments per patient was 2.5 (SD 1.2). However, in six patients renal function did not recover, and they had to be kept on chronic dialysis treatment. Two patients developed fatal complications during their stay in the intensive care unit: a 75 year old man died of gastrointestinal bleeding and an 88 year old woman died of sepsis. Renal function recovered in 34 patients after closely monitored volume administration as dehydration was a common problem. At discharge, creatinine clearance in the patients who were not receiving dialysis was 0.65 (0.27) ml/s, with serum creatinine values between 79 (SD 0.9) μmol/l and 306 μmol/l. Surviving patients were treated with ACE inhibitors (21), β receptor blockers (6), and/or loop diuretics (35). Spironolactone was not reintroduced in our cases.

Discussion

Patients with end stage renal disease in dialysis units often develop hyperkalaemia requiring haemodialysis. More recently the combination of spironolactone plus ACE inhibitors or AT₁ receptor blockers has been found to be associated with severe hyperkalaemia.^6–12 The combination of spironolactone plus ACE inhibitors or AT₁ receptor blockers has been shown to lead to hyperkalaemia, arrhythmia, and death.¹³ In our series, the dose of spironolactone was larger than the 25-50 mg a day used in RALES (the randomised aldactone evaluation study) in most cases. The dosage of spironolactone in RALES was chosen on the basis of this increasing incidence of hyperkalaemia with increasing dosage of spironolactone. This dose limitation had been ignored in our patients and supposedly was an important factor contributing to their hyperkalaemia. Many pharmaceutical companies, however, do not supply the 25 mg dose of generic spironolactone, which might also contribute to overdose in daily practice.

In 19 of the 44 patients serum creatinine concentration was in the range considered safe in RALES—that is, below 221 μmol/l. Only when the creatinine clearance on admission was estimated, was a striking degree of renal failure discovered in many patients. In this situation, estimating the creatinine clearance by using the Cockroft and Gault formula (the SI version)⁵ is very helpful¹⁰: creatinine clearance (ml/s) = (140 - age (years)) × lean body weight (kg)/50 × serum creatinine concentration (μmol/l). For women, the result is multiplied by 0.85.

Renal failure is a well known risk factor for the development of hyperkalaemia after administration of potassium sparing diuretics¹⁴ and is listed in the warnings in the product information. A creatinine clearance of < 0.5 ml/s (present in 34 of our 44 patients) is therefore considered a contraindication to spironolactone (according to product information).

Part of the renal insufficiency in our patients was reversible as on discharge the creatinine clearance improved from a mean of 0.38 (SD 0.22) ml/s to 0.65 (0.27) ml/s except in six patients, who had to remain on long term haemodialysis treatment. Thus, an important part of renal failure was probably prerenal in origin. This interpretation is supported by a recent experimental study that found a powerful natriuretic effect of spironolactone given in combination with an ACE inhibitor.¹⁵ Twenty six of the patients developed muscle weakness and paralysis. Spironolactone intake was reported to be the most common cause of secondary hyperkalaemic paralysis, which often mimics Guillain-Barré syndrome.¹⁶ Thirty five (80%) patients had diabetes mellitus type 2. In this condition aldosterone secretion is low (hyporeninaemic hypoaldosteronism), and this may have contributed to the hyperkalaemia.¹⁷,¹⁸

Conclusion

There seem to be conditions that may lead to the development of severe hyperkalaemia in patients with heart failure who are taking spironolactone and ACE inhibitors or AT₁ receptor blockers: advanced age, dose of spironolactone > 25 mg daily, reduced renal function, and diabetes mellitus type 2. Plasma potassium concentration should be monitored frequently in these patients, in whom we recommend that the dose of spironolactone should be limited to 25 mg a day or even every other day. A test of renal function (at least by applying the Cockroft and Gault formula) before treatment is started is useful as a plasma creatinine concentration < 221 μmol/l does not reliably exclude patients with renal failure. Undetected hyperkalaemia may be suspected as a possible cause of sudden death in some patients treated for heart failure with spironolactone and ACE inhibitors or AT₁ receptor blockers.