Acetazolamide (Diamox; Duramed Pharmaceuticals Inc, Cincinnati, Ohio)is a carbonic anhydrase inhibitor commonly used in ophthalmology to reduceintraocular pressure and to treat some forms of macular edema. Acetazolamide-associatedthrombocytopenia was previously described as part of aplastic anemia or otherorgan involvement; however, evidence that the drug may also cause acceleratedplatelet destruction has never been provided.1,2 Hereinwe report an acetazolamide-induced pure thrombocytopenia with the highestlevel of evidence for a causal relationship of the drug to thrombocytopenia.
A 67-year-old man who had been receiving metformin for 20 years becauseof type 2 diabetes mellitus was admitted to our department for cataract surgery.The day after the surgery, intraocular pressure was elevated and 2 daily 250-mgtablets of acetazolamide were prescribed. As the intraocular pressure normalizedon day 3, acetazolamide was stopped. A routine blood cell count with examinationof the blood smear disclosed an unexpected and isolated thrombocytopenia (plateletcount, 31 × 103/µL) (Figure 1). In the bone marrow, quantitatively and morphologicallynormal megakaryocytes, including some immature forms, were observed, indicatingnormal platelet production. Recovery from thrombocytopenia was shown laterby ambulatory tests. Drug-induced thrombocytopenia was not suspected.
Timeline of platelet count afterfirst introduction of acetazolamide. Vertical lines indicate period of acetazolamideadministration.
Eleven months later, acetazolamide (375 mg/d) was prescribed again,this time for macular edema. Two weeks later, the patient developed extensivepurpura and was seen in our department a week later. He was hospitalized andfound to have a platelet count of 3 × 103/µL (Figure 2). Acetazolamide, but not metformin,treatment was immediately discontinued. The platelet count rose spontaneouslyto 20 × 103/µL the next day, 73 × 103/µL3 days later, and 246 × 103/µL after 10 days. Thrombocytopeniadid not recur.
Timeline of platelet count afterreintroduction of acetazolamide. Vertical lines indicate period of acetazolamideadministration.
To our knowledge, acetazolamide-induced thrombocytopenia has been previouslyreported in only 4 cases.1,3 Moreover,in no case was the causality level high enough to ascertain that the drugwas responsible.
As recommended by the American Society of Hematology,4 diagnosisof isolated thrombocytopenia in elderly patients requires bone marrow aspirationto exclude myelodysplasia. In our patient, no bone marrow disorder was foundto explain the thrombocytopenia.
Diagnosis of acetazolamide-induced thrombocytopenia was based on positivedata including challenge and in vivo rechallenge tests highly suggestive ofcausation: (1) Thrombocytopenia occurred within a few hours or days afteringestion of the drug. (2) Spontaneous recovery from thrombocytopenia wascomplete and sustained after the drug was discontinued, this pattern beingseen on 2 occasions. (3) Reexposure to acetazolamide resulted in recurrentthrombocytopenia.
However, no specific laboratory test was performed to identify circulatingdrug-dependent antiplatelet antibodies and to confirm the diagnosis. Indeed,there are no standard assays for such antibodies, no standardized criteriafor distinguishing positive from negative results, and no data on sensitivityand specificity of these assays based on clinical criteria for a causal relationship.1,5
Although causality assessment methods in pharmacology remain a matterof debate, in our patient acetazolamide caused pure and severe thrombocytopeniawith "certain" evidence according to the World Health Organization systemof causation of a drug reaction,6 with "verylikely" evidence according to the French standardized methodology,7 and with the highest level of evidence ("definite")according to standardized criteria recently developed by Rizvi et al2 (database available at http://moon.ouhsc.edu/jgeorge). Such high levels of evidence for the causal relationship of acetazolamideto thrombocytopenia have never been reported until now, to our knowledge.
Discovery of isolated thrombocytopenia in a patient who is taking severalmedications also presents a challenging clinical problem. The principal interestof the level of evidence is to help clinical decision making about which drugsmay more likely be implicated as a cause of thrombocytopenia and thereforeshould be discontinued as quickly as possible.
Acetazolamide should be considered a definite thrombocytopenia-inducingagent. Potential consequences of thrombocytopenia seem to be limited whenthe drug is prescribed for a few days, whereas it appears different with muchlonger treatment. In that case, regular complete blood cell count, especiallyin the presence of bleeding, should be recommended.5
Correspondence: Dr Renaudier, Department of Hematology, Croix-RousseHospital, 103, grande rue de la Croix-Rousse, Lyon 69004, France (firstname.lastname@example.org).
The authors have no relevant financial interest in this article.
et al. Drug-induced thrombocytopenia: a systematic review of published casereports. Ann Intern Med.
1998;129886- 890PubMedGoogle ScholarCrossref
MW Hematologic reactions to carbonic anhydrase inhibitors. Am J Ophthalmol.
1985;10079- 81PubMedGoogle Scholar
et al. Idiopathic thrombocytopenic purpura: a practice guideline developedby explicit methods for the American Society of Hematology. Blood.
1996;883- 40PubMedGoogle Scholar
G Imputation of the unexpected or toxic effects of drugs: actualizationof the method used in France [in French]. Therapie.
1985;40111- 118PubMedGoogle Scholar