Thrombocytopenia in patients with patent ductus arteriosus after transcatheter closure: perilous complication or harmless incidental finding?

Dear Sir,

Transcatheter closure is now common practice among patients with congenital heart diseases including atrial septum defect, ventricular septum defect, and patent ductus arteriosus (PDA). Though considered safe and effective, it remains plagued by harmful complications such as residual shunt, cardiac tamponade, arrhythmias, and rarely, mechanical haemolysis, while other haematological complications are seldom recorded¹. Here we report a case series of five patients who developed severe thrombocytopenia shortly after transcatheter PDA closure.

Between March 2011 and April 2012, five patients (1 male, 4 fem ales) were reported to have experienced sharp declines in blood platelet counts after transcatheter PDA closure in the Department of Cardiology of our facility, and requested emergency platelet transfusions. The median age of these patients was 23 years (range, 12–42 years). All patients had a PDA with a diameter larger than 5 mm (7–16 mm) and a left-to-right shunt, and three of them also had severe pulmonary hypertension, as determined by right heart catheterisation. After placing PDA occluders, all patients were found to a have a trivial residual shunt. Interestingly, in the three patients in whom transthoracic echocardiography was performed at the 1-month follow-up, the residual shunt was found to have diminished. The cardiovascular characteristics of all the patients are detailed in Table I. Although no major bleeding events other than purpura of the lower limbs were recorded, the thrombocytopenia was very profound. The nadir platelet counts invariably occurred between days 4 and 7 after the transcatheter closure procedure and two patients had platelet counts below 10×10⁹/L while in the other three patients the counts were below 25×10⁹/L. All patients achieved a stable platelet count, ranging from 43×10⁹/L to 92×10⁹/L after day 10, although three patients (patients B, D and E) were given platelet transfusions because their platelet counts dropped below 20×10⁹/L (Figure 1).

Table I.

Cardiovascular characteristics of patients with thrombocytopenia after transcatheter PDA closure.

Patient	Gender/age	PDA type/diameter	PA pressure* (mmHg)	AA pressure* (mmHg)	Residual shunt	Residual shunt at 1 month
A	F/26	funnel/10 mm	49/19/33	134/66/92	(+)	(−)
B	F/42	tubular/7 mm	143/63/73	203/77/84	(+)	(−)
C	F/22	funnel/9 mm	65/39/53	126/64/89	(+)	NT
D	M/12	window/16 mm	105/52/76	131/68/98	(+)	NT
E	F/23	tubular/12 mm	38/20/29	129/59/84	(+)	(−)

Legend PDA: patent ductus arteriosus; PA: pulmonary artery; AA: aortic artery; NT: not tested;

^*PA and AA pressures are shown as systolic/diastolic/mean.

Figure 1.

Platelet counts in the 10 days after transcatheter PDA closure in the five patients studied. The nadir counts occurred between days 4 and 7 after the transcatheter procedure. Two patients had platelet counts lower than 10×10⁹/L while the others had counts less than 25×10⁹/L. All patients achieved a stable platelet count of between 43×10⁹/L and 92×10⁹/L after day 10, although three patients (patients B, D and E) were given platelet transfusions because their platelet counts dropped below 20×10⁹/L.

Thrombocytopenia is a well-recorded complication after invasive procedures; the leading causes include internal bleeding, sepsis, and thrombosis. In the setting of cardiac interventions, both ventricular assist devices and cardiopulmonary bypass are associated with platelet consumption leading to thrombocytopenia, while heparin and glycoprotein IIb/IIIa inhibitors induce thrombocytopenia in an immune-mediated process². In the vessels with high flow velocities and shear stress of the arterial circulation, platelets play the most crucial role in the maintenance of haemostasis due to the “radial dispersion” effect^3,4. All our patients presented with a PDA large enough for there to be a potential risk of a residual shunt after the transcatheter closure procedure and, indeed, a trivial residual shunt was present in all patients after optimised placement of the occluders. Since no other common causes for the low platelet counts, such as heparin-induced thrombocytopenia and sepsis, were evident and apheresis platelets alleviated symptoms remarkably in all three patients who were transfused, we speculated that a local consumption process might be the key underlying mechanism for thrombocytopenia among these patients. Given the endothelial injury and high flow velocity, the site of a residual shunt after PDA closure provides an ideal target for platelet adhesion and subsequent platelet plug formation, which could be a reasonable explanation for the decrease in the residual shunts seen in such a short period in our patients. It is worth mentioning that no life-threatening bleeding occurred in our patients and that the thrombocytopenia, severe as it was, seemed to have been a self-limiting process which did not require aggressive interventions other than platelet transfusion and close monitoring.

In conclusion, this is the first report of severe thrombocytopenia after transcatheter closure with residual shunt among PDA patients. We speculate that local platelet consumption was the major mechanism of the thrombocytopenia, which might be a self-limiting process as the residual shunt diminishes during follow-up.