Observing, explaining and dealing with SOLO/Perceval-mediated thrombocytopaenia Free

Surgical valve replacement is regularly associated with a transient low platelet count in the first postoperative days. Well-established reasons for this common occurrence include cardiopulmonary bypass, haemodilution, hypothermia, blood loss and exposure to exogenous drugs (heparin and protamine). Consequently, this normally affects all valve models and manufacturers almost equally and is therefore not considered unusual. It is, however, unusual and something special when, in addition and unexpectedly, there is an excessive platelet loss after using a certain valve model that cannot be easily explained by the known causes. Now it is no longer a general and assessable concomitant phenomenon but a potentially dangerous side effect. If it refers only to 1 valve prosthesis (or model series), there must be a specific cause that needs to be investigated and understood.

In this issue of the Journal, Lorusso et al. [1] report on sub-analysis results of a prospective randomized trial comparing sutureless and conventional stented valves. They found a significantly greater decrease in postoperative platelet counts in the patients who received the Perceval S sutureless valve prosthesis (formerly Sorin, Saluggia, Italy; now LivaNova PLC, London, UK) compared to the control group (46% vs 32%). The effect was transient and there was no difference in major bleeding and blood loss, transfusion requirements, re-thoracotomy, and stroke events.

The title and, at times, the reading give the impression that this phenomenon could affect all sutureless valves. But to be clear and precise: the more severe early platelet drop is exclusively associated with the Perceval prosthesis investigated in this study. Other sutureless valves have been studied, but this effect was not evident for either the 3f Enable (Medtronic Inc., Minneapolis, MN, USA) [2] or the Intuity Valve (Edwards Lifesciences, Irvine, CA, USA) [3].

The connection with the Perceval has already been reported on [2, 4] and the report by Lorusso et al. is further confirmation of the role of this valve as a strong and independent predictor of increased postoperative thrombocytopaenia. But to get closer to the solution, the Perceval must not be seen in isolation. The Perceval sutureless valve prosthesis was developed from the SOLO stentless valve, and this in turn was a modified version of the Pericarbon Freedom stentless prosthesis. Importantly, there are only differences in the design, but not in the valve tissue used or its processing.

The phenomenon of a severe unexplained temporary decrease in platelets early after implantation of the SOLO valve has been known since 2008 [5]. Since then, this observation has not only been confirmed many times by direct comparison with all available stented valves, but we have also established that this effect occurs exclusively and identically with all 3 valves of the same manufacturer (Pericarbon Freedom, SOLO and Perceval) [2]. They use the identical valve tissue with an exclusive, manufacturer-specific anticalcification treatment. Therefore, the likely key to solving the problem lies in the treatment of the valve tissue. And if this prosthesis tissue can indeed cause thrombocytopaenia then, there is a situation of biomaterial surface–blood interaction here, or more precisely a case of bioincompatibility, which requires clarification. Therefore, the current results by Lorusso et al., while not surprising, are a stark reminder that the reason for this effect is still not widely understood.

Initially, the frequent and unexplained side effect of the otherwise excellent SOLO valve caused many intensive care physicians additional work and long service nights. A large number of tests on diagnosing heparin-induced thrombocytopaenia were carried out, which were regularly negative but ultimately led to the identification of this specific problem [5]. Numerous platelet concentrates were transfused because a high risk of bleeding complications was expected at the time. But then, strangely enough, these did not occur after all, as the report by Lorusso et al. for the Perceval once again confirms. But in fact, this apparent contradiction only adds to the mystery.

We ourselves had introduced and implanted the SOLO prosthesis very early on [6]. After the first reports of associated postoperative thrombocytopaenia, addressing this question was naturally a matter of patient safety. But even after associated bleeding risks could be ruled out, our interdisciplinary group was tempted to pursue this problem further. Our current understanding on the mechanism of SOLO/Perceval-mediated platelet interaction has almost nothing in common with where we started from years ago. More than once we were forced to fundamentally rethink our ideas after conventional assumptions could not be supported by experimental results. We finally came to the following conclusion which appears to us today to be a simple, logical, and conclusive result of our work and that adequately explains all clinical observations.

Several possible causative factors were considered unlikely and, therefore, not further pursued. For example, any causal mechanism related to haemodynamic stress (flow, shear stress) or valve design and implantation technique would lead to persistent, rather than temporary, platelet drops. Furthermore, there is no reason to assume that the superior haemodynamics of the valve with preserved root mobility, particularly large effective orifice areas and low gradients, should be more damaging to platelets than comparable stented valves.

Potential chemical causes include glutaraldehyde or parabens (p-hydroxybenzoic acid esters), but as they are used, respectively, for fixation with all other available biological valves or for storage-solution preservatives without known acute damaging effects on blood cells, they were not considered further. Given that the platelet decrease is temporary, with a nadir on postoperative days 2 and 3, followed by full recovery, persistent toxicity is unlikely. However, to eliminate the last doubts and ambiguities, a possible laboratory platelet measurement error was investigated but could be ruled out [7].

In summary, there was no evidence of either a patient-specific or design-dependent cause for the real and unique effect of more severe postoperative platelet-lowering observed with the Pericarbon Freedom, SOLO and Perceval bioprostheses alike. Rather, it can be logically argued that platelet-lowering effects must result from a specific feature shared exclusively by the 3 bioprostheses.

The most likely candidate is homocysteic acid (HCA), a reactive biochemical compound usually not measurable in peripheral blood, which is uniquely used by the manufacturer for tissue processing to neutralize aldehyde residues. We developed a method to measure HCA levels with a chromatographic (HPLC) protocol using an online pre-column derivatization with o-phtalaldehyde and gradient elution followed by quantification of the fluorescence ratio of HCA to the internal standard norvaline. This technique was then used to determine concentrations of SOLO-associated HCA: (i) in storage solutions of 4 independent valve containers (7.5–7.8 µM), (ii) in homogenates of SOLO valve tissue (0.44 g) in 2 ml of phosphate-buffered saline (5.9–6.2 µM) and in these homogenates incubated with 0.05 N hydrochloric acid to release ionically bound HCA (2.0–2.2 µM), and (iii) in 5 ml of phosphate-buffered saline following agitation (10 rotations/min) with new SOLO prostheses (n = 4) at room temperature for 1 or 24 h (<0.1 µM). In an attempt to reduce patient exposure to HCA, 20 SOLO prostheses were rinsed for 8 min immediately prior to implantation; however, postoperative platelet numbers were identical to those without rinsing.

Although the SOLO valve and its storage solution were found to contain significant amounts of HCA (7.5–8.4 µM), no evidence was obtained to indicate that HCA is released following implantation; therefore, we investigated the effects of direct HCA exposure on platelet activation using platelets from concentrates (n = 9, Fig. 48.9a [6]) and platelet-rich plasma from healthy volunteers (n = 9, Fig. 48.9b [6]). No effect was observed when pieces of SOLO valve tissue (5 mm × 5 mm) were incubated, under rotation, to 2.5 and 5 ml platelet-rich plasma for 1 h followed by exposure to the platelet activators collagen (2 mg/l), arachidonic acid (0.5 M), adenosine diphosphate (10 mM), and thrombin-analogon (40 µM) [8]. Platelets were then directly exposed to various concentrations of L-HCA (H9633 SIGMA-Aldrich, St. Louis, USA) for 30 min followed by stimulation with the same platelet activators. Interestingly, even exposure to extremely high HCA concentrations (in the mM range!) did not inhibit normal platelets (n = 9). However, an effect was observed in smokers with known pre-activated platelets (n = 9, Fig. 48.10b [6]). Therefore, we hypothesized that only previously activated platelets are susceptible to HCA-induced damage. Indeed, platelet activation, which occurs in approximately one-third of platelets during cardiovascular surgery [9], predisposes platelets to HCA-induced damage as a result of conformational changes. The most likely mechanism is via N-methyl-d-aspartic acid (NMDA)-type glutamate receptors, which are massively increased on platelets under conditions of activation. HCA-induced hyperactivation of NMDA receptors has been shown to increase intracellular levels of ionized calcium and reactive oxygen species, followed by progressive degeneration and subsequent cell lysis within 24 h [10]. Non-activated platelets, in contrast, are not affected and remain functionally intact, thereby explaining absence of associated bleeding complications. The HCA-induced decrease in platelets terminates when (i) pre-activated platelets are no longer present and/or (ii) direct contact with platelets is no longer possible as the SOLO valve surface becomes covered with fibrin after only 2 days at the latest, as we were able to observe once with a prematurely explanted prosthesis (picture available from author).

Based on these investigations, transient platelet drops after SOLO/Perceval-implantation can be explained as resulting from selective damage of only ‘pre-activated’ platelets (with upregulated NMDA receptors) upon exposure to homoycsteic acid.

Our investigations and results, more of which can be found elsewhere [6], may help to de-mystify the phenomenon for those interested in it. They support and complement the numerous clinical observations, including now those by Lorusso et al., of the safe clinical use of the Perceval valve prosthesis despite frequently low platelet numbers. Otherwise, if doubt persists, a pragmatic solution would be to limit the implantation to patients with a minimum number of platelets, or to choose an alternative sutureless valve in the first place.

Conflict of interest: none declared.

REFERENCES