Elise Huisman

The focus of this thesis is on children with platelet disorders (PD). The aim is to improve the diagnostic algorithms identifying the causes and underlying mechanisms of PD in order to improve counseling and treatment of children suffering from a PD. Until recently, Glanzmann Thrombasthenia and Bernard-Soulier syndrome were the only clearly identifiable PD subgroups based on routinely available platelet function testing. The last years, the role of molecular genetic testing has increasingly become part of the diagnostic work up. This is important, as PD are heterogenous with regard to genetic background and clinical expression, but also associated with symptoms in various organs resulting in additional comorbidities. The basis of this work is the DoReMi cohort study, which was initiated to better define PD and advance the knowledge of PD in children. This study aims to move from the oversimplification of PD as a disease with either a shortage (thrombocytopenia) or a dysfunction of platelets (thrombocytopathy) towards a rare hereditary bleeding disorder classification based on complex underlying pathways. A scientific boost to better understand the multicolored spectrum of PD was therefore urgently needed. The studies in this thesis contribute to emergence of evidence-based differentiation of PD (Chapter 1). In Part I of this thesis, we evaluate and review the current clinical and diagnostic land scape identifying where improvements can be made. In Part II, we investigated current dilemmas regarding therapeutic options for PD. PART I In Chapter 2, we present a scoping review on health-related quality of life (HRQoL) of children with PD and other rare bleeding disorders. We found that research investigat ing HRQoL in children with PD and their caregivers is lacking. Only one publication on this topic was identified, but used a questionnaire that was developed for children with immune mediated thrombocytopenia, an acquired PD. Extrapolation of disease-specific questionnaires applied in other rare bleeding disorders, such as the extensively studied hemophilia, is clearly not adequate for use in the multi-colored PD disease-entity, as female-related bleeding is not represented. We also show that when HRQoL is investigated with questionnaires, only reporting the mean or median HRQoL of the study population is not sufficiently informative. We pro pose that future HRQoL research in children with bleeding disorders, such as PD, should focus on the outliers, as the children and caregivers scoring high on HRQoL as well as the those scoring low may guide clinicians to recognize valuable or ineffective coping skills or points to disease and treatment related factors that influence HRQoL positively or negatively. Furthermore, pooling of children with various disease severity should be avoided. In Chapter 3, we describe the clinical symptoms and laboratory parameters in a cohort of 98 children with PD (DoReMi study). We found that 40% of children with PD reported other comorbidities. Overall, children with PD report mainly mucocutaneuous and postoperative bleeding. Interestingly, only a small proportion reports the presence of petechiae. Sex-related differences were found in bleeding presentation and underlying etiology of the PD. In boys, the differential diagnosis of bleeding after circumcision should include PD. In girls, heavy menstrual bleeding occurred frequently and was often the presenting symptom of their PD. Besides sex-related differences, different subgroups within PD were identified. We confirmed that PD combined with specific coagulation factor deficiencies occurred mainly in patients with Noonan syndrome. Based on our findings, we propose a classification with two main groups. These are based on the results of molecular genetics and functional analysis. When a molecular genetic cause is identified, we propose to use the term inherited PD (IPD). When no such molecu lar genetic cause is identified to explain the platelet function disorder leading to either a low platelet count (thrombocytopenia) or a platelet dysfunction (thrombocytopathy), the PD should be classified as a functional PD (FPD) Within these two main groups of IPD and FPD, a subgroup can be identified. For IPD, we identify mutations that lead to isolated disorders in the hematopoietic tract. This so-called isolated PD (ISPD) will lead to a PD and may be accompanied by other red or white blood cell disorders with this not necessarily so. Or the PD is part of an underlying syndrome, a syndrome-associated PD (SAPD). Further specification can be given by adding the gene related to or the syndrome associated with the PD, e.g. Noonan-associated IPD. For the second group, the FPD, the subgroup classification is based on the low platelet count (quantitative disorder), or functional defect (qualitative disorder), that can be fur ther specified by adding the functional defect, e.g. COX-mediated FPD. In Chapter 4, we investigated if mass-spectrometry (MS) analysis of the platelet proteome is able to better diagnose PD. As proteins are the functional derivatives of our genetic code, analyses of the platelet proteome may be more informative to identify the genetic disorders underlying IPD and SAPD. Moreover, these results may also help to unravel the modifying pathways in this complex disorder with a heterogeneous presentation. To test this hypothesis, the platelet proteome of 80 children and family members with a PD and an identified genetic defect was analyzed. We demonstrated that >4000 proteins are detectable in these platelets using the applied set-up and MS technique. These proteins can be grouped by either their location or functional role, such as ‘granule-associated proteins’ and ‘hemostasis-involved proteins’. Variability and reproducibility was good or acceptable for almost all proteins. One important limitation of the technique is that gene products involved in early megakaryopoiesis, like FLI1p and ETV6p, are not detectable in the peripheral platelets. In Chapter 5, we investigated the clinical phenotype and laboratory results of 14 patients with a variety of 11q aberrations. Partial deletions or duplications of the long arm of chromosome 11 (11q), ranging from 1-20 Mb, are rare and considered a contiguous gene deletion or duplication syndrome. Terminal 11q deletions occur most frequently. Break points arising within or distal to 11q23.3 with a deletion extending to the telomere lead to a clinical entity known as Jacobsen syndrome. The clinical phenotype of Jacobsen syn drome includes increased bleeding tendency due to low platelet counts and/or platelet dysfunction. Typically, giant alpha granules may be observed known as Paris-Trousseau syndrome, but abnormal dense granules and microkaryocytes have also been reported. Besides Jacobsen syndrome, interstitial deletions as well as interstitial and terminal dupli cations of 11q are also reported. Detailed data on bleeding phenotype, platelet numbers and platelet function in non-terminal 11q disorders is lacking. We observed that patients with interstitial 11q deletions as well as terminal 11q dupli cations have a similar phenotype as patients with terminal 11q deletions (Jacobsen syndrome). Most frequently reported bleeding symptoms were hematomas, bleeding from small wounds, perioperative bleeding and heavy menstrual bleeding in females. We demonstrated that despite normal platelet number, platelet dysfunction was highly prevalent. Therefore, when 11q aberrations are present, platelet count and function must be tested by light transmission aggregometry (LTA). Remarkably, the abnormal responses to thrombin receptor-activating peptide (TRAP) that we observed may reflect impaired effect of thrombin in vivo. This may point to storage pool disease (SPD), a subgroup of PD with abnormal platelet granule function. This was supported in our report by abnormal ATP-ADP ratio and abundance of NUBP2p and ATP5Sp observed by MS-based analysis of the platelet proteome. Finally, abnormal alpha granules were seen by electron micros copy in two patients with an interstitial and a terminal 11q deletion. There is evidence from other PD, like Hermansky-Pudlak syndrome type 2 or Grey platelet syndrome that granule function in the neutrophils may also be abnormal. As patients with 11q disorders, including Jacobsen syndrome, are reported to be susceptible to bacterial and prolonged viral and fungal infections, we investigated the function of neutrophils as well as of B and T cells in Chapter 6. We found that in 30% of patients’ granulocyte function was abnormal, but this observation did not lead to a distinct clinical phenotype fitting granulocyte disorders. Importantly, T cell dysfunction was present in almost all patients. This supports the findings of a high burden of infections in patients, consisting of respiratory, ear-nose-throat infections or secondary bacterial infections of varicella skin lesions, despite normal immunoglobulin G levels. A large proportion of patients also reported chronic eczema, warts, and chronic fungal infections. Furthermore, decreased numbers of B cells (57%), as well as T cells (36%) and abnormal T cell function (86%) were demonstrated in patients with 11q aberrations except for interstitial 11q duplications. We did not find a clear phenotype-genotype correlation with FLI1 and ETS1 involvement in the majority of patients. Therefore, we concluded that other candidate genes on 11q may play a role in both platelet and immune dysregulation. Fundamental research is needed to further understand the role of these genes coding for hematopoietic transcription fac tors, like FLI1. In Chapter 7 we investigated the role of the transcription factor FLI1, a member of the erythroblast transformation specific (ETS) family, on megakaryopoiesis and platelet production in a patient with a de novo heterozygous nonsense mutation in FLI1 (NM_002017.5(FLI1): c.297del, p(Met100*)). Effect of this in potential truncating variant on megakaryocyte maturation was investigated by comparing wild-type (WT) FLI1p and FLI1 p.Met100* in Meg01 cells on RNA and proteome level. By doing so, we demonstrated that the truncated FLI1 p.Met100* proteins that were expressed in Meg01 cells, upregu lated both megakaryoblasts and erythroblasts in the progenitor stage of differentiation, by advantaging KLF1 impact. This led to a distinct pattern on flow cytometry with sig nificantly elevated and prolonged expression of CD235a on erythroid cells and CD36 on megakaryocytes, as well as elevated KLF1p expression by FLI1Met100*. However, after this megakaryocyte progenitor phase, maturation and differentiation to megakaryocytes were both accelerated and impaired in the FLI1Met100* Meg01 cells. This was not seen in WT FLI1, as up-regulation of WT FLI1 alone only led to accelerated differentiation but normal maturation. Our findings therefore suggest that FLI1 p.Met100* impairs megakaryocyte maturation through a dominant-negative manner contributing to the congenital throm bocytopenia. Potentially this is the result of upregulated erythroid features at the expense of megakaryocytic development. In Chapter 8, we present two siblings with pathogenic variants in another transcription factor gene that tightly regulates megakaryopoiesis, the ETV6 (formerly TEL), which is also a member of the ETS family. Pathogenic variants in ETV6 lead to autosomal dominant thrombocytopenia. ETV6 acts mainly as a transcriptional repressor, forming dimers through its pointed (PNT) domain, interacting with cofactors via its central domain, and binding DNA through its conserved ETS domain. It is in this domain that most pathogenic variants are found. In the two patients with an inherited thrombocytopenia due to a class 4 variant in ETV6 (NM_001987.4(ETV6): c.1105C>T, p.(Arg369Trp, in short ETV6-R369W), we found that 424 genes were upregulated and 203 were downregulated in comparison to identical genes in the wild-type (WT) cells. This is consistent with ETV6’s repressive function in normal conditions. Products from erythroid genes (GYPA, CD36, and KLF1) were reduced, while products from genes associated with hematopoietic lineage and platelet activation pathways (CD44, KIT, GP6, and FLI1) were increased. Overexpression of ETV6-R369W in wild-type Meg01 cells produced a similar pattern, suggesting a dom- inant-negative mechanism. In CD34+ cultures, ETV6-R369W accelerated megakaryocyte differentiation. CD44 was also elevated in both patient’s platelets, suggesting its potential as a biomarker for ETV6-related thrombocytopenia. PART II Platelet transfusions are indicated in case of acute and/ or severe bleeding (therapeutic transfusion) or to prevent bleeding (prophylactic transfusion) to resolve thrombocytopenia or thrombocytopathy. The development of HLA or HPA antibodies can of course complicate treatment, as platelets will not increase sufficiently, therefore forming a potentially life-threatening complication. We demonstrate in Chapter 9 that prophylactic matching for HLA is not only regarded as good clinical practice in literature reviews but, according to the survey performed, is also already widely applied by a majority of Dutch clinicians treating PD patients. Literature and guidelines reviewed unfortunately lack essential information on non-transfusion and transfusion related risk factors contributing to HLA or HPA antibody formation, to estimate incidence estimation. We conclude that regarding HLA antibody prevention, prevention of HLA formation is the key leading to preservation of random platelet transfusions in acute setting for patients with Glanzmann Thrombasthenia and Bernard-Soulier syndrome. This benefit outweighs the effort of timely-ordered and more costly HLA matched platelets in elective setting, until more knowledge is gained on risk groups and incidence of HLA antibody formation. For HPA antibody prevention, we demonstrated that screening for HPA antibodies may be helpful in very specific clinical management and risk assessment, such as pregnancy. Assessment can be based on a risk model based on HPA genotyping. For pregnant mothers with Glanzmann Thrombasthenia or Bernard-Soulier syndrome, we recommend fetal monitoring during pregnancy when maternal HLA, HPA or isoantibodies are present, as well as postnatal platelet count measurements in the neonate. In Chapter 10, we present a systematic review to provide the essential insights on the association between thrombocytopenia or prophylactic platelet transfusions on major bleedings in premature neonates. This is important information to establish what platelet counts in premature neonates triggers prophylactic platelet transfusions. We screened 8734 publications and included 36 publications that met established study inclusion criteria. Of these 36 publications, only seven were of sufficient scientific quality to include in the final analysis. The other 29 were excluded due to inexact information on timeframe between major bleeding and platelet count or major bleeding and platelet transfusion. This so-called temporal ambiguity, led to cause-and-effect relationships influencing the outcome. In the seven papers, n=3 addressed the association between prophylactic platelet transfusions and bleeding, and n=4 platelet count and bleeding (n=4). No association was seen between prophylactic platelet transfusions or platelet count with major bleeding in n=5, but in n=2, an associated between thrombocytopenia and pulmonary hemorrhage was seen. Overall, we concluded that the causal relationship between thrombocytopenia, platelet transfusions and major bleeding in extreme premature neonates can neither be proven nor rejected in literature. Therefore, the clinical need for prophylactic platelet transfusions in neonates with an asymptomatic thrombocytopenia could not be resolved. Chapter 11 describes the use of desmopressin (1-deamino-8-D-arginine vasopressin) in children with PD. Desmopressin enhances primary hemostasis by activating alpha granules and cyclic adenosine monophosphate, that in return leads to exocytosis of von Willebrand Factor (VWF) in the vascular endothelium. Mode of action is derived from its use in von Willebrand disease (VWD) and mild hemophilia A patients. Previously, only one double-blind randomized crossover trial in 18 PD patients was performed, besides few case-series describing small numbers of (pooled) adult and pediatric PD patients. In most reports, desmopressin response was quantified by reduction of bleeding time, a test that is no longer applied. We, therefore, conducted an observational cohort study to evaluate desmopressin response in 51 children, who were aged between 4-18 years, and had various subtypes of PD. We evaluated repeated measurements of von Willebrand factor (VWF activity) and Platelet Function Analyzer (PFA) results. We defined the response as ‘good, partial or ‘poor’ based on a combination of VWF rise and/or PFA normalization. We found that desmopressin response was ‘good’ in 30, ‘partial’ in 13 and ‘poor’ in eight of these children. We concluded that not all PD patients respond adequately to desmopressin when the response is quantified by rise in VWF activity or PFA normalization, and that this this was especially true for children classified as inherited PD (IPD) or Syndrome-associated PD (SAPD). Lastly in Chapter 12, the most relevant outcomes of the studies are placed in a broader perspective. We discuss the diagnostic and therapeutic approaches that we have inves tigated and describe the knowledge gaps identified by our studies, suggesting future research aims.