Evidence review: safety

Advice

The summary of product characteristics for Florinef (100 microgram tablets; fludrocortisone acetate) reports that fludrocortisone is contraindicated for people with systemic infections, unless specific anti-infective therapy is used, and in people with hypersensitivity to any of the ingredients.

Suppression of the inflammatory response and immune function by fludrocortisone increases susceptibility to infections and their severity; chicken pox, shingles and measles are of particular concern. Patients should be advised to avoid exposure to these diseases and seek medical advice without delay if exposure occurs. The summary of product characteristics reports that because fludrocortisone is a potent mineralocorticoid, dosage and salt intake should be monitored to avoid hypertension, oedema or weight gain, and that monitoring of serum electrolyte levels is advisable during prolonged therapy ('prolonged' not further defined).

The summary of product characteristics reports that adrenal cortical atrophy develops with prolonged use and may persist for years after stopping treatment. It says that withdrawal after prolonged use must be gradual to avoid acute adrenal insufficiency and fludrocortisone should be tapered off over weeks or months depending on the treatment dose and duration.

The summary of product characteristics reports that people should be warned that potentially severe psychiatric adverse reactions may occur with systemic steroids. Symptoms typically emerge within a few days or weeks of starting the treatment. The summary of product characteristics reports that when adverse effects occur with fludrocortisone, they are usually reversible after stopping treatment and that the incidence of predictable adverse effects is dependent on the dosage, frequency and duration of treatment. It says that people receiving fludrocortisone should have close monitoring for adverse effects associated with corticosteroid therapy. The common adverse effects of systemic corticosteroids may be associated with more serious consequences in older adults.

In Campbell et al. (1975), there was a statistically significant increase in supine systolic blood pressure (SBP) with fludrocortisone compared with placebo. However in Rowe et al. (2001), mean supine SBP was similar between fludrocortisone and placebo, and it was reported that no participants had a change in SBP of more than 40 mmHg while receiving treatment. In Schoffer et al. (2007), mean supine SBP with fludrocortisone was similar to the mean supine SBP before treatment.

In Campbell et al. (1975), 2 people, both with known intermittent proteinuria and low plasma albumin, developed pitting ankle oedema within 1 week of starting fludrocortisone; 1 of these people also reported frontal headache and breathlessness. Both participants were reported to have completed the 3‑week intervention period, and the adverse effects were said to rapidly subside after completing treatment. The other 3 participants who completed the study are reported not to have experienced any side effects while taking fludrocortisone.

There was an increase in mean body weight (mean difference 2.7 kg, p<0.001) and total plasma volume (mean difference 329 ml, p<0.02) with fludrocortisone compared with placebo. There was a statistically significant rise in serum sodium (142.6 mmol/litre [±2.2] versus 140.6 mmol/litre [±1.3], p<0.02) and a statistically significant reduction in serum potassium (3.6 mmol/litre [±0.4] versus 4.3 mmol/litre [±0.2], p<0.01) with fludrocortisone compared with placebo. There were no significant differences between fludrocortisone and placebo for urinary electrolyte concentrations, plasma and urine osmolality and creatinine clearance (no p values given).

Schoffer et al. (2007) found that withdrawal from the study was lower in the fludrocortisone group (1 of 17 [5.9%]) than the domperidone group (3 of 17 [17.6%]). These withdrawals all occurred within the 1st week of treatment. Six people reported side effects while receiving fludrocortisone (including 2 reports of nausea and single reports of chest discomfort, morning headache, light-headedness and dizziness) compared with 5 people receiving domperidone (2 reports of nausea and single reports of chest pain, abdominal pain, palpitations and headache); statistical significance of between-group comparison not reported. On the basis of unused medication in returned pill bottles, adherence to treatments was reported as excellent, with an average missed dosage of 1 tablet of fludrocortisone and 3 tablets of domperidone.

Rowe et al. (2001) found that withdrawal from the study was higher in the fludrocortisone group (13 of 50 [26%]) than in the placebo group (8 of 50 [16%]); statistical analysis not reported. In the fludrocortisone group, 4 people withdrew because they developed depression, 2 for worsening symptoms, 2 for abdominal discomfort, 1 for worsening headache, 1 who was randomised but did not receive treatment because of major depression, and 1 because of unrelated medical illness. In the placebo group, 3 people dropped out because of no improvement, 1 for panic symptoms and tachycardia, 1 for increased fatigue and 1 stopped taking the medication because of severe light-headedness, fatigue and sweating. One person from each group withdrew because of hypertension and 1 person withdrew from each group because of unwillingness to continue with the study.

At least 1 adverse effect was reported during treatment by 61% of people receiving fludrocortisone compared with 71% receiving placebo; data for these adverse events are not reported. No important adverse effects were reported by the study authors as lasting long after treatment was stopped. Mean weight increased by 1.1 kg in the fludrocortisone group and by 1.2 kg in the placebo group. Mean serum sodium concentration was higher during the treatment period in the fludrocortisone group (141.9 mEq/litre [±2.0]) compared with the placebo group (140.3 mEq/litre [±2.3], p=0.003).

Limited data on longer-term safety with fludrocortisone are available from a prospective observational study in 64 adults (mean age 80 years, range 58–98 years) with 1 or more hypotensive disorders with an average follow-up of 12 months (range 2–21 months) (Hussain et al. 1996). There were inconsistencies in the reporting of the doses of fludrocortisone between the abstract and in the full text of the publication, which were corrected in a later edition of the journal (Hussain et al. 1996; erratum). Of 64 people, 13 died of causes unrelated to hypotension. Adverse events were reported in 38 of 64 people (59.4%) with 17 stopping treatment.

There were inconsistencies in the reporting of reasons for drug withdrawal between figures reported in the abstract and figures reported in the table. Figures for drug withdrawal reported here are taken from the table, because it contains other additional information about adverse events. Reasons for drug withdrawal were: cardiac failure (n=7), systolic hypertension (n=4), depression (n=3), no benefit (n=2) and stroke (n=1). Hypokalaemia developed in 8 participants but there were no withdrawals for this reason. However, the observational nature of this study, its mixed population, with people with postural hypotension (n=17), vasodepressor carotid sinus syndrome (n=19), mixed vasodepressor carotid sinus syndrome and postural hypotension (n=14) and other hypotensive disorders (n=14), and the lack of a control arm limit the conclusions that can be drawn.