Iloprost-induced thrombocytopenia: a case proven by rechallenge

To the Editors:

A 63-yr-old female underwent aortic and mitral valve replacement surgery (ATS 18 and 29 mm, respectively; ATS Medical, Inc., Minneapolis, MN, USA) for aortic stenosis and mitral incompetence in August 2009. Pre-operatively, she was in New York Heart Association class III and exhibited comorbid hypertension, obstructive sleep apnoea (managed with night-time continuous positive airway pressure), obesity (body mass index of 42 kg·m⁻²) and hypercholesterolaemia. Pre-operative right and left heart catheter investigation demonstrated a normal left ventricular ejection fraction, mean aortic valve gradient of 42 mmHg (normal 0–10 mmHg), valve opening area of 0.7 cm² (normal >2 cm²), mean pulmonary arterial pressure of 48 mmHg (normal 12–16 mmHg), pulmonary vascular resistance of 225 dyn·s·cm⁻⁵ (normal 100–200 dyn·s·cm⁻⁵), pulmonary capillary wedge pressure of 34 mmHg and cardiac output of 4.97 L·min⁻¹. Her medication included an angiotensin-converting enzyme inhibitor, diuretics and aspirin.

Her post-operative course was complicated by a sternal wound infection, mediastinitis, pneumonia and renal failure requiring continuous high-flow haemodiafiltration. On the 14th post-operative day, she developed right heart failure (central venous pressure (CVP) of 25 mmHg, mean arterial pressure (P̄_a) of 60 mmHg and severely impaired right ventricular function on echocardiography) due to sepsis- and pneumonia-related exacerbation of her initially mild post-capillary pulmonary hypertension. She required mechanical ventilation in a spontaneously breathing mode and circulatory support with noradrenalin and dobutamine. Treatment with inhaled iloprost (Ilomedin®; Bayer Schering Pharma, Berlin, Germany) was commenced in order to reduce pulmonary arterial pressures and right ventricular work. A total of 10 μg iloprost were administered by nebulisation via a Micro-Cirrus™ device (Intersurgical, Wokingham, UK) every 3 h (80 μg·24 h⁻¹), with a resulting increase in P̄_a from 60 to 80 mmHg and reductions in serum lactate level from 10 to 1.3 mM and CVP from 25 to 16 mmHg during the subsequent 48 h. This treatment with iloprost was off-label in terms of indication and mode of application. On day 1 of iloprost treatment, the patient's platelet count was 295×10⁹ cells·L⁻¹, but fell continuously to 25×10⁹ cells·L⁻¹ on day 6, when iloprost was withdrawn (fig. 1). Concomitant medication included vancomycin, heparin, dobutamine, noradrenalin, fluconazole, ceftriaxone, remifentanil, haloperidol and morphine. Although screening for anti-heparin platelet factor 4 antibodies was negative, heparin was discontinued. The platelet count began to recover as soon as iloprost was withdrawn. Falling P̄_a and rising CVP and serum lactate levels, however, prompted recommencement of iloprost (120 μg·24 h⁻¹). The platelet count fell again on this rechallenge, and iloprost was again withdrawn. Heparin, however, was recommenced and platelet counts continued to recover (fig. 1). There was no bleeding and platelet transfusion was not required. Disseminated intravascular coagulation was excluded by the presence of normal fibrinogen and D-dimer levels. Pseudothrombocytopenia was excluded by the absence of platelet clumps in the blood film. Additionally, in the peripheral blood, there was no evidence of global bone marrow suppression or thrombotic microangiopathy.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1–

Platelet count over time during iloprost treatment (: lower limit of normal platelet count). Concomitant therapy with potential thrombocytopenic agents is shown.

The differential diagnosis of the thrombocytopenia included heparin-induced thrombocytopenia type II, disseminated intravascular coagulation, haemofiltration-induced thrombocytopenia, pseudothrombocytopenia, and iloprost-, vancomycin- or dobutamine-induced thrombocytopenia. The first three causes could be excluded on the basis of normal findings in their respective diagnostic tests, and there was no temporal relationship between either vancomycin or dobutamine exposure (both previously described to be associated with thrombocytopenia 1, 2) and thrombocytopenia. Thus thrombocytopenia caused by iloprost was the likeliest cause.

In 1998, George et al. 3 established the criteria for the diagnosis of drug-induced thrombocytopenia. These are: 1) the candidate drug preceded thrombocytopenia and recovery from the thrombocytopenia was complete and sustained after the drug was discontinued; 2) the candidate drug was the only drug used prior to the onset of thrombocytopenia or other drugs were continued or reintroduced following discontinuation of the candidate drug with a sustained normal platelet count; 3) other aetiologies of thrombocytopenia were excluded; and 4) re-exposure to the candidate drug resulted in recurrent thrombocytopenia. In the present case, iloprost was the only medication that fulfilled all four criteria and can thus be classified as a definite cause of the drug-induced thrombocytopenia.

Iloprost (Ilomedin® or Ventavis®; Bayer Schering Pharma) is a synthetic prostacyclin analogue used in the treatment of pulmonary arterial hypertension. Commonly reported side-effects of iloprost are hypotension, cough, headache, gastrointestinal disturbance, chest pain, dyspnoea, bronchospasm and wheezing 4. Although iloprost is a potent platelet inhibitor, thrombocytopenia has not previously been reported in the medical literature. Post-marketing surveillance data regarding Ventavis® revealed cases of gingival bleeding and epistaxis within 1 month of commencing therapy 5; however there are no data on platelet count or the frequency of this complication. Pharmacovigilance database enquiry (Swissmedic national database on spontaneous reports of suspected adverse drug reactions and World Health Organization VigiBase^TM 6 accessed in December 2009) revealed five cases of possible iloprost-induced thrombocytopenia in four countries during 2004–2009. In two cases, the route of administration was inhaled, in two intravenous and in one unknown. In three cases, recovery of platelet counts and no attributable deaths were reported. Epoprostenol, another prostacyclin analogue, however, has been associated with the development of thrombocytopenia 7.

The exact mechanism of this iloprost-induced thrombocytopenia remains unclear, but the rapidity of onset would be compatible with antibody-mediated platelet destruction. It is possible that iloprost or its metabolites dinoriloprost and tetranoriloprost form complexes with pre-existing immunoglobulin (Ig) molecules and that these then bind to prostacyclin receptors on platelets, thereby causing Fc-mediated platelet activation and subsequent destruction through elimination via the reticulo-endothelial system, similar to the mechanism whereby vancomycin causes platelet destruction 1. The slower recovery of the platelet count following the second exposure to iloprost seen in the case described here would support this hypothesis. Confirmation of this mechanism would require detection of iloprost-dependent IgM and/or platelet-reactive IgG in serum.

• ©2010 ERS