Topical dorzolamide and metabolic acidosis in a neonate

We describe a neonate with bilateral Peter’s anomaly who became unwell and developed a metabolic acidosis after commencing topical dorzolamide. He was fully investigated to exclude other causes of acidosis, and subsequently improved on discontinuation of topical treatment. To the best of our knowledge, there have been no reports of topical carbonic anhydrase inhibitors causing metabolic acidosis in children or adults.

A 5 day old boy was referred to a tertiary paediatric ophthalmology unit with bilateral corneal opacities for consideration of penetrating keratoplasty. He had a normal Apgar score at delivery at 35 weeks’ gestation and weight 2.3 kg. In addition, he had had full screening investigations including blood gases, abdominal ultrasound, and DMSA scans because of a prenatal history of intrauterine growth retardation with suspicion of a single kidney.

Ocular examination revealed total left corneal opacification and a small opacity of the right cornea inferiorly. Intraocular pressures measured were normal but since digitally the eyes felt firm and applanation tonometry is unreliable in thinned corneas, he was commenced on Trusopt (MSD) eye drops, three times daily to both eyes. He was to be reviewed 2 weeks later.

Seven days following the commencement of Trusopt at his regular paediatric follow up appointment, he was found to be sleepy with poor feeding and poor capillary refill. There was no history of diarrhoea or vomiting. Arterial blood gases revealed a metabolic acidosis with a pH 7.08, Pco₂ 4.2 kp, Po₂ 11.3 kp and bicarbonate 9.3 mmol/l and base excess of −20.2. There were no markers of infection with negative blood, urine, stool, throat, and nasal cultures. Anion gap, serum electrolytes, liver function and urinalysis for pH, specific gravity, and electrolytes were also unremarkable. Renal ultrasound and DMSA scan showed a normal functioning single right kidney.

As the cause for the metabolic acidosis at this stage was unknown he was given intravenous cefotaxime, flucloxacillin, half correction bicarbonate infusion followed by oral sodium bicarbonate supplements for 3 days. He showed some improvement with treatment; however, he remained significantly acidotic and unwell. At routine ophthalmic review 5 days later, while free of all other treatments, the eye drops were stopped and he showed spontaneous next-day resolution of his acidosis. He symptomatically improved and gained weight over the subsequent few days (Fig 1).

Figure 1.

Trend in metabolic acidosis over time.

Topical dorzolamide has been shown to cause significant reduction in intraocular pressure (IOP) in children and is well tolerated.¹ Secondary glaucoma is well recognised in cases of Peter’s anomaly and raised IOP is well known to cause corneal clouding.² Congenital corneal opacities necessitate urgent treatment in order to reduce amblyopia, and therefore it is essential to exclude glaucoma. Topical Trusopt (MSD) is used routinely at the department of ophthalmology, Great Ormond Street, as it is thought to have lower potential for adverse systemic effects than topical β blockers.

Topical dorzolamide is a potent inhibitor of CA-II¹ and this inhibition decreases the rate of aqueous humour secretion consequently lowering IOP. In the proximal renal tubule CA-II is also required to sustain maximal rates of HCO₃ reabsorption. Significant systemic inhibition of carbonic anhydrase has not been observed and there has been an absence of demonstrable metabolic effects in adults.^2–6 However, with the oral carbonic anhydrase inhibitor, acetazolamide, the renal carbonic anhydrase involvement and acidosis have been shown to be proportionally related to the plasma concentration levels of the drug. The dose per kg systemic absorption of topically administered dorzolamide would be expected to be higher in neonates/infants of lower body weight compared with adults.

Metabolic acidosis with normal anion gap and serum electrolytes in the absence of diarrhoea, as in this case, is more likely to be due to proximal renal tubular bicarbonate loss. Spontaneous improvement of the acidosis on termination of the topical dorzolamide is strongly suggestive of the culpability of dorzolamide. It is unclear as to why this happened, but factors such as prematurity, low birth weight, renal tubular immaturity, and one functioning kidney may have led to poor handling of drug elimination at a higher systemic concentration. Although we feel dorzolamide is a relatively safe topical antihypertensive treatment, this case underlines the need for caution when treating neonates.