What is the clinical spectrum of Thrombocytopenia?
Thrombocytopenia is a hematologic pathology in which counts of platelets (thrombocytes), that intervene in blood clotting and wound healing, are below the lower limit of normal, 150x109/L for adults. Thrombocytopenic patients are at increased risk of bleeding, thrombosis, organ damage, and mortality.
Thrombocytopenia clinical presentation, prevalence, degree of morbidity, and fatality rate varies according to disease subtype 1, 2. The most common cases found in intensive care units are related to drug-induced thrombocytopenia, thrombocytopenic thrombotic purpura, hemolytic uremic syndrome, and disseminated intravascular coagulation.
Several conditions share the pathophysiology of decreased platelet production by megakaryocyte progenitors or increased platelet destruction 1, 3. Decreased production was reported in aplastic anemia bone marrow failure, drug-related and cancer-related bone marrow suppression, viral infections, systemic conditions such as nutrient deficiency and sepsis, and inherited thrombocytopenia. Increased destruction is present in primary immune and drug-induced thrombocytopenia, lymphoproliferative disorders, autoimmune conditions like systemic lupus erythematosus, in chronic infections, as well as in several non-immune related conditions.
Aplastic anemia (AA) is driven by pancytopenia - deficiency of all blood cells, causing fatigue, severe bleeding and vulnerability to life-threatening infections 3, 4. Drug-induced thrombocytopenia (DIT) presents with a sudden onset of petechiae and purpura, mucosal bleeding, gastrointestinal bleeding, and in rare cases intracranial hemorrhage 5, 6. Chemotherapy-induced thrombocytopenia (CIT) is a frequent complication of cancer therapies, resulting in increased risk of bleeding, dose reduction, and treatment discontinuation negatively affecting oncologic outcomes 7, 8. Heparin-induced thrombocytopenia (HIT) can cause arterial and venal thrombosis, leading to stroke or pulmonary embolism associated with a mortality rate of up to 20% 1, 2, 9. Radiation-induced thrombocytopenia (RIT) typically presents with reduced platelet count and mild (petechiae, bruising, mucosal bleeding) to moderate (gastrointestinal, hematuria) bleeding symptoms, that may require interruption of radiotherapy in patients with malignant diseases 21, 23. Primary immune thrombocytopenia (ITP) and secondary ITP are characterized by low platelet count, petechiae, bruising and mild to severe hemorrhagic episodes 1, 2, 10. Primary inherited thrombocytopenia (IT), such as Gray platelet syndrome, Wiskott-Aldrich syndrome and Alport syndrome, and secondary IT, such as Shwachman–Diamond syndrome and congenital amegakaryocytic thrombocytopenia (CAMT) present with a variable bleeding risk 3, 11. Additionally, patients with IT are at risk of developing other life-threatening disorders. For instance, an increased risk of malignancies and autoimmunity was noted for individuals with Wiskott-Aldrich syndrome, while CAMT may present with severe bone marrow aplasia in infancy 12. Thrombotic thrombocytopenic purpura (TTP) is a rapidly progressive thrombotic microangiopathy characterized by low platelet count, fever, hemolytic anemia, neurological symptoms and renal dysfunction. Widespread microvascular thrombosis causing multiorgan ischemia and failure is the main driver of mortality in TTP. If left untreated, TTP has a mortality of about 90% 13, 14. Disseminated intravascular coagulation (DIC) exhibits both excessive clot formation (microvascular thrombosis) and bleeding, resulting in mortality in up to 50% of patients 1, 15, 16. Symptoms of DIC also include bruising, low blood pressure, shortness of breath, and confusion. Similarly to TTP, the main driver of mortality in DIC is microvascular thrombosis that can produce multiple organ failures, predominantly acute kidney injury and acute respiratory distress syndrome. The symptoms of hemolytic uremic syndrome (HUS) include low platelet counts, hemolytic anemia, neurological complications, and acute kidney failure 1, 17, 18.
What do we know about the etiology of Thrombocytopenia?
The causes of thrombocytopenia are heterogenous and can be internal, external, or a combination of both. These typically result in bone marrow dysfunction-caused decreased production, immune-mediated increased destruction, coagulopathy-related increased consumption, and splenic pooling sequestration.
Aplastic anemia (AA)
AA is an autoimmune disorder that leads to growth inhibition and apoptosis of hematopoietic stem and progenitor cell (HSPC), including megakaryocyte progenitors, resulting in pancytopenia 4. The etiology of AA can be idiopathic (believed to be immune-mediated), hereditary (Fanconi anemia, Dyskeratosis congenita), toxic exposure (chemicals, drugs, radiation) or viral infection (Epstein-Barr virus, cytomegalovirus, hepatitis viruses) 3, 19. Acquired idiopathic AA, virus-associated AA and sometimes drug-associated AA are driven by CD8 cytotoxic T cell–mediated autoimmune attack against HSPC, initiated and sustained by IFN-γ and TNF-α-secreting CD4 T cells 57. Triggering events of immune activation in AA include virus or drug-initiated aberrant antigen presentation, genetic predisposition to immune dysregulation and loss of immune tolerance.
Drug-induced thrombocytopenia (HIT) - Chemotherapy-induced thrombocytopenia (CIT) - Heparin-induced thrombocytopenia (HIT)
Several drugs are associated with DIT, through a variety of mechanisms, including immune mediation (antibiotics - quinine, vancomycin, sulfonamides; anticonvulsants – carbamazepine, phenytoin), and hematopoietic progenitors DNA synthesis inhibition (antivirals - ganciclovir) 3. Multiple distinct mechanisms are believed to underly drug-induced immune thrombocytopenia pathogenesis, including drug-dependent antibody, drug-specific antibody and hapten-dependent antibody, the dynamics of which are not fully understood 5, 6, 20. In the first case, drug-dependent antibodies bind platelet glycoproteins only in the presence of the drug, leading to rapid platelet clearance upon re-exposure of the sensitized individual to drug.
CIT is commonly caused by cytotoxic effects of chemotherapy, damaging HSPC 7. Several mechanisms of CIT have been proposed, such as direct suppression of megakaryocyte progenitors in the bone marrow, alteration of the bone marrow microenvironment (stromal cells, cytokines) that supports hematopoiesis, and apoptosis-triggering DNA damage 8. Beyond megakaryocyte suppression, other mechanisms that may underly CIT involve immune-mediated platelet destruction (fludarabine, oxaliplatin) and splenic sequestration.
The widely used anticoagulant heparin can trigger an abnormal immune response in which the body produces specific antibodies against platelet factor 4 (PF4)-heparin complexes, resulting in HIT 1, 2. Secondary to binding of anti-PF4-heparin antibodies to platelet FcγRIIA receptors, platelets are activated - change shape, increase surface expression of adhesion molecules and release procoagulant microparticles, promoting thrombin generation. Despite the fact that HIT antibodies are drug-triggered, they are not strictly drug-dependent and can in some cases cross react with PF4 alone, producing delayed-onset HIT after heparin is discontinued 9.
Radiation-induced thrombocytopenia (RIT)
RIT can result from medical (radiotherapy) and non-medical (occupational, military, environmental) exposure to radiation. While both short-term high-dose or long-term low-dose radiation can cause damage to the hematopoietic system, the nature and severity of the damage depends heavily on the exposure type, dosage rate and duration 21, 23. In contrast to chemotherapy, severe marrow injury in radiotherapy occurs only within the irradiated volume and damage to the medullary stroma may be irreversible in cases of high radiation doses 21. The underlying mechanisms of RIT remain to be clarified and are believed to involve direct (apoptosis of HSC, DNA damage) and indirect processes (damage to stromal cells, vascular injury, inflammatory cytokines) 23.
Primary and secondary immune thrombocytopenia (ITP)
Primary ITP is an autoimmune disorder with no identifiable external trigger 1, 2, 3, 10. In primary and secondary ITP, platelet destruction involves both arms of the adaptive immune system - humoral (mediated by B cells) and cellular (mediated by CD8 cytotoxic T cells). Platelets are opsonized by autoantibodies produced by B cells against platelet glycoproteins, and subsequently cleared by macrophages via their antibody-binding Fcy surface receptors. An imbalance between FcyRIIA (activating) and FcyRIIB (inhibitory) receptors promotes excessive antibody-dependent platelet destruction by macrophages. Measurement of Th9 and Th17 cells in active ITP patients and ITP patients in remission indicated that T helper cells were elevated in disease state, promoting B cell proliferation and class-switch recombination. This enhanced T helper cell-mediated B cell activation leads to increased autoantibody production and greater platelet destruction 24.
Secondary ITP typically has an identifiable external cause - drugs, infections, malignancies like chronic lymphocytic leukemia, and autoimmune conditions like systemic lupus erythematosus 1, 3. Viral infections, such as HIV, hepatitis C, and cytomegalovirus infections, were found to precede the onset of ITP symptoms in certain cases. Due to molecular mimicry between viral antigens and platelet surface proteins, autoreactive B cells may produce autoantibodies against platelet glycoproteins like GPIIb/IIIa, marking them for destruction by macrophages. Apart from antibody-dependent cellular phagocytosis, other immune-mediated mechanisms that may underly platelet destruction include natural-killer cells dysfunction, complement-dependent cytotoxicity, and CD8 T cell-mediated cytotoxicity.
Primary and secondary inherited thrombocytopenia (IT)
Most forms of primary IT are caused by rare mutations in genes implicated in megakaryocyte proliferation, differentiation and proplatelet formation, although their exact mechanisms remain yet to be unveiled 3, 12. To date, several dozen genetic variants with complete or incomplete penetrance have been implicated in primary IT, among which MYH9 (cytoskeletal motor protein - platelet formation), WAS (cytoplasmic signaling protein - platelet morphology), ANKRD26 (transcription factor - megakaryocyte proliferation and differentiation regulation), RUNX1 (transcription factor - HSC differentiation regulation), and NBEAL2 (vesicle-trafficking protein - alpha granule biogenesis). For example, in Wiskott-Aldrich syndrome, mutations in WAS protein result in impaired proplatelet formation and abnormally small platelets that are cleared by the reticuloendothelial system. Common variants in genes like GP6 (glycoprotein VI receptor - platelet aggregation), ITGA2B (integrin αIIb protein - platelet aggregation), and SH2B3 (cytokine signaling adaptor protein - HSC proliferation) may act as modifiers of disease severity, contributing to inter-individual differences in IT phenotype and pathophysiology.
In secondary IT, mutations commonly affect broader bone marrow function, leading to defective HSC function, ribosome biogenesis, or cell cycle regulation. Pathogenic variants associated with secondary inherited thrombocytopenia, such as SBDS (ribosome biogenesis protein - HSC function), DNAJC21 (co-chaperone protein - ribosomal maturation), and EFL1 (GTPase - ribosomal maturation), are mostly inherited in an autosomal recessive manner 11, 12. For instance, congenital amegakaryocytic thrombocytopenia is a rare autosomal recessive disorder with loss of function mutations in MPL gene encoding thrombopoietin receptor which is essential for the commitment differentiation of multipotent HSC to megakaryocytes, resulting in low platelet count from birth 25, 26.
Thrombotic thrombocytopenic purpura (TTP)
About 95% of TTP cases are immune TTP (iTTP) acquired via an autoimmune mechanisms against ADAMTS13, a liver-produced plasma von Willebrand factor (VWF)-cleaving protease, in most cases by an unknown trigger. ADAMTS13 is a metalloprotease that cleaves the highly pro-thrombotic ultra-large von Willebrand factor (UL-VWF) multimers into smaller, less adhesive fragments. Without ADAMTS13, persistently adhesive UL-VWF multimers bind platelets, initiating microthrombus formation and driving continuous platelet aggregation into these thrombi, resulting in depletion of circulating platelets 2, 13, 14. Antiplatelet (antiaggregant) drugs, immunosuppressive agents, HIV, estrogen-containing birth control pills, and pregnancy are the most commonly listed triggers for ADAMTS13 autoantibody formation causing iTTP.
Congenital thrombotic thrombocytopenic purpura (cTTP), also known as Upshaw–Schulman syndrome, is caused by biallelic homozygous or compound heterozygous mutations in the ADAMTS13 gene, producing deficiency in ADAMTS13 activity. Some cTTP mutation types are associated with an early onset severe disease form while others may enable residual enzyme activity, resulting in milder disease symptoms and later onset.
Disseminated intravascular coagulation (DIC)
DIC is secondary to an underlying condition, mainly sepsis, trauma, cancer, infections and obstetric complications 1, 15, 16. These conditions produce inflammatory triggers that cause massive release of procoagulants (tissue factor, cytokines) from endothelial cells and monocytes, leading to excess thrombin generation within the vasculature, consumption of platelet, fibrinogen, and clotting factors, and consumption coagulopathy that impairs the organism's ability to form new clots where needed. In hyperfibrinolytic DIC, that may occur in response to clotting, the excessive activity of plasmin leads to premature degradation of fibrin clots, preventing stable clot formation and facilitating uncontrolled bleeding. The innate immune response mediated by monocytes and neutrophils at the sites of tissue injury or infection plays a central role in development of DIC, including through damage-associated molecular patterns that are released from injured cells, pathogen-associated molecular patterns that originate from infections, and proinflammatory cytokines released by tumor cells 15. The pathways activated vary according to the trigger type (cancer, trauma, infection), explaining why some DIC patients present with predominant thrombosis while others present with massive bleeding.
Hemolytic uremic syndrome (HUS)
Typical HUS is caused by Shiga toxin-producing E. coli (STEC) strains 1, 17, 18. Shiga toxin is a ribosome-inactivating protein that inhibits translation, leading to cell stress, apoptosis, and inflammation. The two major Shiga toxin families - Stx1 and Stx2 - bind preferentially to glycolipid globotriaocylceramide (Gb3) receptors, which are highly expressed on glomerular endothelial cells and podocytes. By releasing tissue factor, cytokines, VWF, collagen and other procoagulants, the Shiga toxin-damaged endothelial cells enable platelet aggregation and fibrin deposition, forming microvascular thrombi that consume circulating platelets, causing thrombocytopenia. Microthrombi that accumulate in glomerular capillaries lead to renal ischemia and acute kidney injury. As thrombi narrow the kidney microvasculature, the passing red blood cells are mechanically sheared, causing microangiopathic hemolytic anemia. Accounting for approximately 5-10% of all HUS cases, atypical HUS is caused by autoimmune disorders, genetic mutations, and malignancies.
How similar are human and animal skeletal systems?
This is not an exhaustive list of species-specific differences, nor can one be made given their unknown full extent, but rather an example of how these differences impact the face, construct, and predictive validity of animal models.
Not all species-specific differences can be accounted for in animal models, as there are hundreds of them, their relevance to thrombocytopenia is unclear, and their interaction with other organ systems in the animal model makes it difficult to predict how they would have behaved within the human system.
There are significant differences between mouse and human platelet production and regulation that negatively affect researcher's ability to mimic thrombocytopenia phenotypes, to understand its pathomechanisms, and to develop effective therapies.
Species-specific differences in immune system
Innate and adaptive immunity both play a central role in a great number of thrombocytopenia types. Platelets are actively monitored by immune cells for signs of senescence, dysfunction, or opsonization. Macrophages, dendritic cells and neutrophils carrying the Fcy receptor (FcyR) recognize platelet-bound antibodies, inducing phagocytosis of platelets. Cytokines IFN-γ, IL-6 and TNF-α-secreting CD4 T cells initiate CD8 cytotoxic T cell–mediated apoptosis of HSPC and alter the bone marrow microenvironment that supports hematopoiesis. IgG/IgM autoantibodies bind to platelet surface glycoproteins, activating the classical complement pathway that prompts phagocytosis of opsonized platelets by macrophages. Upon binding of anti-PF4-heparin antibodies to platelet FcγR, platelets are activated, ultimately resulting in generation of vascular thrombi. T helper cells promote B cell activation, autoantibody production, and greater platelet destruction, just to mention a few examples 2, 8, 10, 15, 19. It follows that inter-species differences in immune system produce divergences between animal models and humans in susceptibility, disease phenotype, pathophysiology, and responses to treatments for thrombocytopenia.
Extensive differences between humans and mice were demonstrated in the structure of innate and adaptive immunity, including in balance of leukocyte subsets, defensins, Toll receptors, inducible NO synthase, the NK inhibitory receptor families Ly49 and KIR, FcR, Ig subsets, the B cell (BLNK, Btk, and λ5) and T cell (ZAP70 and common γ-chain) signaling pathway components, Thy-1, γδ T cells, cytokines and cytokine receptors, Th1/Th2 differentiation, costimulatory molecule expression and function, Ag-presenting function of endothelial cells, and chemokine and chemokine receptor expression 27. In addition, it was shown that transcriptional responses to acute inflammatory stresses of different etiologies in mouse models do not correlate with human acute inflammatory diseases 28. Major species-specific differences were also found in transcriptional regulation, chromatin state and higher order chromatin organization. Notably, cis-regulatory sequences next to immune-system-related genes have shown the most divergence, as they have apparently undergone more rapid evolution than others 29.
For instance, mice lack the genetic equivalent of FcγRIIA activating receptor for IgG, which in humans plays a key role in recognition of IgG antibody-coated platelets by macrophages. Moreover, mouse monocytes and macrophages express a greater number of inhibitory FcyRIIB receptors than their human counterparts, limiting anti-body dependent platelet clearance 30, 31. As a result, mouse models of platelet-mediated immune activation may underestimate human risk and produce misleading results for monoclonal antibody therapies.
It was suggested that using animal models with a humanized immune system might improve translatability to humans, however, such an approach would face persistent challenges: the cross-talk between the human immune system and the rest of human organ systems, as well as inter-patient variation in polymorphisms and expression of FcR, HLA Class I/II and other inflammatory pathway genes cannot be recapitulated in animals. Genotype analysis conducted in patients with immune thrombocytopenia (ITP) indicated that these polymorphisms were correlated with disease susceptibility and therapeutic response, explaining variable success in clinical trials 31, 32, 33.
It is also important to highlight that species-specific differences in immune system contribute to the poor negative predictivity of preclinical animal testing, exposing patients to immune-mediated drug adverse effects, including thrombocytopenia 34.
Species-specific differences in platelet production, clearance and turnover rates
The liver-produced hormone thrombopoietin (TPO) is the primary physiological regulator of platelet production. By binding to the MPL receptor on HSC, megakaryocyte progenitors and mature megakaryocytes, TPO promotes HSC differentiation toward megakaryocyte lineage, proliferation of megakaryocyte progenitors and cytoplasmic expansion of megakaryocytes, enabling the extension of proplatelets that fragment into platelets. In drug-induced liver injury (DILI), impaired liver function can result in reduction of TPO levels, leading to decreased platelet synthesis.
Mice have faster platelet turnover rates compared to humans, supported by higher circulating TPO levels and faster clearance rate of TPO compared to humans 35. The accelerated platelet production may confer a higher resistance to certain thrombocytopenia-inducing conditions seen in humans.
Deleterious mutations in TPO receptor MPL cause congenital amegakaryocytic thrombocytopenia (CAMT) in humans 25, 26. In contrast, mice with double knockout of TPO and its receptor MPL maintain normal life expectancy and do not develop aplastic anemia or bone marrow failure, pointing to species-specific hematopoietic regulation 36.
Species-specific differences in platelet receptors
Platelet activation for clot formation is triggered via G-protein coupled protease-activated receptors (PAR) located on the surface of platelets. In humans, thrombin, a serine protease, activates PAR1 and PAR4 by cleaving their extracellular domain. However, platelets in mice, rats, guinea pigs, hamsters, and rabbits, do not possess PAR1. Instead they express PAR3 which, unlike PAR1, does not in itself mediate thrombin signaling but acts as a cofactor in tandem with PAR4 37. As a result, rodent platelets may respond differently to thrombin than human platelets. Moreover, the absence of PAR1 in rodents makes them poor models for developing and testing PAR1-targeting antithrombotic drugs like Vorapaxar 38.
The hormone TPO mediates its effect on platelet production through binding to the megakaryocyte progenitor transmembrane cell receptor MPL (myeloproliferative leukemia protein) 39. Histidine at position 499 (H499) in the juxtamembrane domain of MPL, that stabilizes MPL in an inactive conformation, is specific to humans and chimpanzees 40. In all other species (mice, rats, rabbits, dogs, non-chimpanzee primates etc.), Mpl has a higher propensity for dimerization than human MPL, which in turn leads to enhanced activation of JAK2/STAT5 and other downstream pathways that drive platelet production. As a result, model organisms may appear less sensitive to human disease-inducing MPL mutations, leading to erroneous hypotheses on thrombocytopenia mechanisms and overestimation of drug efficacy in preclinical testing.
Other examples of human features that contribute to species-specific differences in MPL structure and activity are threonine at position 496 (T496) and tryptophan at position 491 (W491). Indicated for treatment of primary immune thrombocytopenia and aplastic anemia, the TPO receptor agonist eltrombopag binds to W491, promoting MPL dimerization and subsequent activation of platelet production. In commonly used model organisms (mice, rats, rabbits, dogs, baboons, rhesus macaques), W491 is replaced by glutamine (Q491) or other amino acids, rendering eltrombopag inactive 41. The example of eltrombopag is a cautionary tale highlighting how species-specific differences may mislead drug development.
Species-specific differences in platelet central signaling cascade
Exogenous factors such as vascular injury, drugs, inflammation, and infection can trigger a coordinated and tightly regulated response of platelet molecular networks known as central signaling cascade. Systems biological pathway analysis-aided multi-omics comparison of central platelet signaling cascades in mice and humans, shows major differences in expression levels and regulatory fine-tuning between the two species that may adversely affect the ability of mouse models of thrombocytopenia to recapitulate human-relevant pathophysiology and reliably predict human responses to treatments 42.
For instance, human central interactors PLCB2 (phospholipase C beta 2 - Ca signaling and platelet activation via G protein coupled receptors-GPCR pathways), MMP9 (metalloproteinase 9 - inflammation and tissue remodeling), BDNF (brain-derived neurotrophic factor - neuron plasticity, platelet storage and release), ITPR3 (inositol 1,4,5-trisphosphate receptor type 3 - intracellular calcium release for platelet activation), and SLC25A6 (mitochondrial ADP/ATP translocase - cell metabolism) were absent in murine transcriptome and proteome datasets, whereas murine central interactors GNG12 (G protein subunit gamma 12 - GPCR signaling), PRKCE (protein kinase C epsilon - signal transduction, cytoskeletal dynamics, and platelet function), and ADCY9 (adenylate cyclase 9 -cAMP production influencing platelet inhibition) were absent in human datasets.
The expectation that researchers should account for these inter-species differences is unrealistic. Vertical, temporal and functional integration of multi-omics data from a single species is in itself challenging. Even the most sophisticated AI models struggle to predict the effects of multi-layer molecular, cellular and physiological divergences between species, especially in complex biological pathways like platelet signaling. To add a supplementary layer of complexity, platelet signaling cascade pathways involve non-linear species-specific feedback loops, molecular/cellular interactions and alternative compensatory pathways that are virtually impossible to process through AI without overfitting and oversimplification, resulting in high risk of misinterpretation of data when applied across species.
Species-specific differences in platelet aggregation factors
Wild type rodents do not develop thrombotic thrombocytopenic purpura (TTP), and in ADAMTS13 knock-out mice UL VWF do not accumulate to the same pathological extent as in humans 77, 78. This human-specific susceptibility to TTP can be explained by inter-species differences in VWF structure (protein sequence and multimer size), regulation (threshold for VWF release), and clearance mechanisms of VWF (hepatic, enzymatic) that are not well understood 43.
In hemolytic uremic syndrome (HUS), E. Coli-produced Shiga toxins enter the cell by binding to Gb3 receptors via sites on Shiga toxin’s pentameric B subunits. There are major inter-species differences in cellular Gb3 expression and localization, hindering development of effective treatments for HUS 44. Gb3 is widely distributed in human tissues, including in platelets, heart, kidney, lung, liver, smooth muscle, epithelium of gastrointestinal tract, and neurons. While Gb3 is abundantly expressed in human glomerular endothelial cells, podocytes, mesangial cells, and tubular epithelia, in murine kidneys it is expressed primarily in the proximal tubular epithelial cells. In contrast to humans, rodents develop gastrointestinal and renal tubular epithelial lesions but fail to develop glomerular endothelial damage 45.
Species-specific differences in genetics and gene expression regulation
Researchers that perform experiments in mice often argue that the mouse genome is similar to the human genome. This statement is, however, inaccurate and misleading. Even if over 90% of human protein-coding genes have a counterpart in mice, the gene orthologs can have a different structure and function across species 46, 47, 48. Additionally, the remaining up to 10% human-specific protein-coding genes may play a crucial role in the human disease.
While human and mouse protein coding genes may have 85% of DNA sequence in common, this is only on average – meaning that some genes may be highly conserved (up to 99%) while others may diverge significantly (as low as 60%) 49. Even if a given human protein-coding gene shared 99% of nucleotide sequence with its mouse ortholog, a single nucleotide difference can result in an amino acid substitution that may dramatically alter the protein’s structure and function 50.
Furthermore, protein-coding DNA represents just 1-2% of the human genome. 98-99% of DNA is composed of non-protein-coding elements, a number of which play a vital role in gene expression regulation 51. Notably, about 90% of disease-associated variants identified by GWAS reside in non-coding regions of the genome 52.
For instance, mice lack a direct equivalent of the human 5’ UTR non-coding regulatory region of the ankyrin repeat domain 26 (ANKRD26) gene, making it very difficult to model ANKRD26-related thrombocytopenia phenotype and pathophysiology in mice. In humans, point mutations in the 5′ UTR of the ANKRD26 gene lead to persistent expression of ANKRD26 transcription factor in HSPC which causes MAPK pathway hyperactivation, impairing platelet production. However, the mouse Ankrd26 gene has a different 5′ UTR sequence and the upstream binding motifs for transcription factors RUNX1 and FLI1 are absent in mice, meaning that even if mutations are introduced into the mouse 5′ UTR, they do not disrupt the same regulatory mechanisms as in humans 11, 53, 54.
Several other mutually interactive human organ systems influence the function of the bone marrow and can affect susceptibility to thrombocytopenia, including endocrine system (hypothyroidism, steroid-responsive thrombocytopenia, cirrhosis-induced low platelet counts), digestive system (gut microbiome-related inflammation and nutrient malabsorption) and cardiovascular system (congestive heart failure-related platelets splenic sequestration, hypotension-impaired bone marrow perfusion).
Face validity - How well do animal models replicate the human disease phenotype?
Attempts to model human thrombocytopenia in animals have had a varying degree of success over the last six decades of research, depending on the disease (sub)type modeled, the species used, and the method of experimental induction.
Aplastic anemia (AA)
A variety of experimental induction methods was employed to model aplastic anemia (AA) in animals, including injection of allogeneic lymphocytes to mimic autoimmune bone marrow destruction, genetic engineering to replicate hereditary AA types, infection with viruses linked to AA in humans, and chemical exposure to agents like benzene and busulfan to trigger toxic injury to HSC. In benzene-induced hematopoietic toxicity mouse model, BALB/c mice with damage in bone marrow, spleen and thymus showed pancytopenia and immunosuppression 55. The effects of viruses, such as lymphocytic choriomeningitis virus (LCMV), were found to be variable and lacking significant bone marrow cell loss in mice 56. Induction of AA through infusion of allogeneic lymphocytes after irradiation or thymectomy effectively recapitulated bone marrow failure and pancytopenia in mice, although without capturing the chronic and progressive nature of human AA 58, 59.
Drug-induced thrombocytopenia (DIT) - Chemotherapy-induced thrombocytopenia (CIT) - Heparin-induced thrombocytopenia (HIT)
To model immune-mediated drug-dependent DIT, which is caused by antibodies that bind platelets only in the presence of the drug, humanized mouse models like NOD/scid mice were infused with human platelets, drugs and patient-derived drug-dependent antibodies (DDAb) 62. While these models reproduced antibody-mediated platelet destruction for a few well characterized drugs, they cannot determine whether and how a drug will trigger DDAb formation in a naïve host.
Injected intraperitoneally, the cytotoxic chemotherapy agent fluorouracil (5-FU), that is indicated in treatment of several types of cancer, induced profound myelosuppression in mice 60. To enhance the severity of thrombocytopenia, a combined chemotherapy-radiation mouse model was developed 61 in which mice were treated with multicycle regimen of carboplatin plus irradiation, resulting in anemia and thrombocytopenia.
Transgenic mice expressing human PF4 and FcγRIIA replicated binding of anti-PF4/heparin antibodies, platelet activation, thrombocytopenia and thrombosis 64. However, since animal models of HIT mostly used passive transfer of patient-derived antibodies, they did not capture the autoimmune-like delayed onset HIT that in patients cause significant morbidity and mortality 9.
Radiation-induced thrombocytopenia (RIT)
When exposed to total body radiation, model organisms (rodents, dogs, non-human primates) presented with key hallmarks of RIT - bone marrow damage and reduced platelet production, that correlated with radiation dose and exposure duration 23. Inter-strain differences in response of mice to total body radiation were noted, with BALB/c being most sensitive and the C57BL/6 most resistant, complicating comparison of drug efficacy 22. A major limitation of assessing the effect of radiation in rodents is that smaller and thinner rodent bodies allow for more uniform radiation penetration than in humans, leading to overestimation or underestimation of bone marrow suppression severity and platelet recovery kinetics 23.
Primary and secondary immune thrombocytopenia (ITP)
Rodent models of ITP were developed through passive transfer of anti-platelet antiserum or platelet-specific monoclonal antibodies induced acute thrombocytopenia 68. Unlike in ITP patients, mucosal bleeding and petechiae were rarely observed in mouse models of ITP, limiting their utility for studying clinical bleeding risk. In contrast to human ITP which is often chronic-relapsing, rodent models did not capture the long-term immune dysregulation. While the innate response phase was present in mouse models of ITP generated by passive transfer of an alloantibody, no evidence of adaptive immunity was found. Proinflammatory cytokines and chemokines that play a critical role in autoimmune inflammation in human ITP, including interleukin-1α/2/6/7/23, granulocyte-macrophage–colony-stimulating factor, monocyte chemoattractant protein, macrophage inflammatory protein, RANTES, tumor necrosis factor-α, and interferon-γ, remained at negligible levels after ITP induction 66. Differing from mice, dogs that developed ITP naturally produced anti-platelet antibodies, exhibiting chronic thrombocytopenia, bleeding, and petechiae 67.
Primary and secondary inherited thrombocytopenia (IT)
In gray platelet syndrome (GPS), mutations in human NBEAL2 gene result in deficiency of platelet α-granules (disrupted biogenesis of α-granules that carry VWF, fibrinogen, PF4 and other proteins involved in clotting), and defective formation of functional platelets (reduced platelet count, abnormally large platelets, dysfunctional platelets), leading to increased risk of bleeding and bone marrow fibrosis 69. In rodent models of GPS, both similarities and differences to human GPS were reported. For instance several Nbeal2-null mouse strains 70, 71 recapitulated deficiency of alpha granules biogenesis, abnormally large platelets, reduced platelet count and defective platelet aggregation, but without bone marrow fibrosis frequently present in GPS patient biopsies. Contrarily to human GPS, characterized by quantitative deficiency in alpha granules and qualitative defect in their content, Wistar Furth rats did not exhibit reduced levels of α-granules per platelet - instead the granules had reduced levels of clotting proteins 72. Nbeal2-null C57BL/6 × 129S5 70 and Nbeal2-null C57BL/6J 71 mice also showed inter-strain differences in proplatelet formation, obscuring the understanding of GPS pathophysiology.
Individuals with MYH9 related disease (MYH9-RD) lack non-muscle myosin heavy chain IIA cytoskeletal protein that is critical for proplatelet formation, resulting in large platelets, low platelet count and dysfunctional platelets, and subsequent increased risk of heavy and prolonged bleeding 73. Several mouse model of MYH9-RD were developed by genetic editing. Homozygous Myh9 mice did not survive as embryos and heterozygous Myh9 mice showed only mild thrombocytopenia without major bleeding 74.
In humans with ANKRD26-related thrombocytopenia, point mutations in the 5′ UTR of the ANKRD26 gene result in persistent expression of ANKRD26 transcription factor in HSPC, which causes MAPK pathway hyperactivation and impairs platelet production, leading to mild to moderate thrombocytopenia and increased risk of hematologic malignancy 12, 53. However, experimental inactivation of ANKRD26 gene in mice failed to produce symptoms of thrombocytopenia, highlighting species-specific differences in gene function 54.
Caused by biallelic mutations in the MPL gene coding for platelet TPO receptor, congenital amegakaryocytic thrombocytopenia (CAMT) presents with absent megakaryocytes and severe thrombocytopenia from birth, which progresses to bone marrow failure in early childhood 25, 26. Mice engineered by gene knock-out to be Mpl-deficient or TPO-deficient showed reduced numbers of megakaryocytes and severe thrombocytopenia. However, in contrast to CAMT patients, these mice had normal levels of erythrocytes and leukocytes in their peripheral blood and did not develop bone marrow failure during their lifetime 75, 76.
Thrombotic thrombocytopenic purpura (TTP)
In an attempt to model the congenital thrombotic thrombocytopenic purpura (cTTP), ADAMTS13 knock-out mice were generated. In contrast to humans with cTTP who often spontaneously develop TTP symptoms early in life, ADAMTS13-deficient mice did not show any TTP symptoms, possibly due to presence of alternative VWF-cleaving proteases in mice that can compensate for ADAMTS13 deficiency 81. Features of acute TTP with severe thrombocytopenia were visible only after injection with Shiga toxin as trigger and only in some ADAMTS13-deficient mice 77. Administration of recombinant human VWF containing UL‐VWF multimers to ADAMTS13-deficient mice induced TTP in all treated mice, although brain and kidney lesions typical of human TTP were not reported 78.
To experimentally induce acquired immune TTP (iTTP), a combination of two inhibitory monoclonal antibodies against murine ADAMTS13 was injected in mice. Despite the accumulation of UL-VWF multimers, TTP-like symptoms did not occur in this mouse model of iTTP and could only be induced when recombinant human VWF was injected as an additional trigger 79. In search for a more human-relevant iTTP model, inhibitory murine anti-human ADAMTS13 monoclonal antibody was administered intravenously to baboons 80. This baboon model presented with severe thrombocytopenia and microangiopathic hemolytic anemia with VWF-rich microthrombi in most organs, without the need for an additional trigger. In contrast to humans with TTP, baboons did not develop renal and neurologic dysfunction and none died during the study, possibly owing to species-specific differences in vascular resilience. Due to persistent immune dysregulation in human iTTP (autoreactive memory B cells, defective regulatory T cells), relapse is relatively frequent in iTTP patients once treatment is stopped, requiring long-term monitoring and management. However, since animal models of iTTP were obtained by passive antibody transfer, they do not recapitulate the relapsing-remitting disease course.
Disseminated intravascular coagulation (DIC)
Numerous attempts have been made to reproduce DIC-like symptoms in mice, rats, dogs, non-human primates and other animal species by employing a variety of experimental induction methods, including blood loss, hemocoagulating poisons, viral infection, tumour inoculation, thermal injury, and infusion of lipopolysaccharides (LPS), tissue factor, cytokines, and activated factor Xa-phospholipids. While the International Society on Thrombosis and Haemostasis provides clinical criteria for diagnosing DIC, these criteria were rarely used to benchmark animal models of DIC, leading to generation of clinically irrelevant animal models of DIC that are not comparable to human DIC 82. The lack of human relevance, high disease heterogeneity, and species-specific differences all contribute to poor comparability across animal models and limited predictive value for human outcomes, may it be for animal models of DIC or animal models of other human diseases. While coagulation activation, platelet consumption, and microvascular thrombosis were recapitulated in most animal models of acute DIC, multiple organ failures seen in human DIC were often absent or mild. With the exception of tissue factor and activated factor Xa-phospholipid infusion, most DIC in vivo models tended to overrepresent thrombotic features and underrepresent fibrinolysis - bleeding diathesis seen in acute promyelocytic leukemia patients 83, 84, 85.
Hemolytic uremic syndrome (HUS)
The most common approach of experimental induction of HUS in animals consists of inoculating purified Shiga toxin (Stx1/Stx2) or Stx-producing E. Coli (STEC) strains, often in combination with LPS to amplify the inflammatory response. The resulting HUS mouse models recapitulated key hallmarks of HUS - thrombocytopenia, hemolytic anemia, renal microvascular thrombosis and glomerular injury 86. Gnotobiotic piglets infected with STEC strains only partially mimicked HUS since they developed neurological complications but without severe hemolytic anemia, thrombocytopenia and renal failure 87. Non-human primates showed less susceptibility to developing HUS from STEC infections than humans. Bonnet macaques injected with STEC did not develop thrombocytopenia nor hemolytic anemia 88, while baboons showed renal injury, thrombocytopenia and anemia only when injected intravenously with purified Stx1. Contrarily to what is known about HUS progression in humans, Stx–induced thrombotic coagulopathy in baboons was found to originate in the kidney and hematologic features, that are thought to precede renal injury in humans, appeared to arise as a consequence of localized renal microvascular damage 89.
Construct validity - How well do the mechanisms of disease induction in animals reflect the currently understood etiology of the human disease?
Thrombocytopenia has a multitude of etiologies, sometimes with no identifiable trigger. However, animal models often fail to replicate the complexity of human thrombocytopenia and its mechanisms. The above described inter-species differences in immune system, genetics, platelet receptor, aggregation and central signaling cascade play an important part in the inability of animal model to faithfully replicate the pathogenesis of various human thrombocytopenia subtypes.
Aplastic anemia (AA)
The use of environmental pollutants and alkylating agents in model organisms to investigate its effects on hematopoiesis is consistent with what is known about the real world exposure to these chemicals in humans 55. Benzene and busulfan were found to cause direct toxicity to human HSC, although the molecular mechanisms of response to toxicants remain unclear and may differ between species. Species-specific differences in tropism for human viruses and in immune responses may explain variable results of viral infection-induced mouse models of AA, that largely failed to produce full blown AA 56. Induction of AA through infusion of allogeneic lymphocytes after irradiation or thymectomy effectively recapitulated T cell–mediated destruction of HSPC 58, 59. Nevertheless, irradiation and thymectomy are nor representative of human AA etiology and this acute immune-mediated mouse models may not replicate chronic immune dysregulation seen in humans. Human AA often develops gradually and persists for months, meaning that this AA induction method hinders investigation of the mechanisms of disease progression. Patient studies suggest that AA is antigen-driven 57, however, animal research into AA did not permit to identify specific antigens responsible for the initiation of immune attack and eventual relapse after immunosuppressive therapy, stalling progress toward antigen-specific therapies. Although the roles of IFN-γ and TNF-α in immune-mediated mouse models of AA have been identified, animal models of AA fall short of uncovering the precise immune mechanisms that initiate and sustain AA in humans.
Drug-induced thrombocytopenia (DIT) - Chemotherapy-induced thrombocytopenia (CIT) - Heparin-induced thrombocytopenia (HIT)
Since they rely on pre-existing antibodies from patients, immune-mediated DIT mouse models do not simulate the initiation phase of the immune response 62. NOD/scid mice are severely immunodeficient and not representative of the human immune regulation. Immune-mediated DIT mouse models are therefore inappropriate for studying the mechanisms of human immune responses and for predicting drug-dependent immunogenicity in humans. Moreover, these models do not recapitulate the complexity of human drug-induced immune thrombocytopenia and the diversity of its pathologic mechanisms 20.
Animal models of CIT (rodents, non-human primates, dogs) replicated bone marrow suppression and impaired megakaryopoiesis observed in CIT patients 60, 61. Yet, several therapies developed in animal models of CIT, such as TPO analogues and TPO receptor agonists, failed to produce desired results in cancer patients 63, pointing to inter-species differences in mechanisms of CIT and platelet recovery. Another concern is that CIT animal models do not capture the complexity of human CIT, that may involve immune dysregulation, genetic variants (drug metabolism) and comorbidities (anemia, infection).
Even though mouse models of HIT were partially humanized to better reflect human-specific features of PF4 and FcγRIIA 64, genetic editing cannot eliminate the totality of above described inter-species differences, may they be in genetic background, platelet PAR1 thrombin receptor, platelet signaling cascade or immune regulation. These human-specific features play an important role in crosstalk between immune and coagulation pathways and in thrombin feedback, influencing responses to antithrombotic therapies. In addition, most HIT animal models use passive transfer of patient-derived antibodies or immunization with PF4/heparin complexes, meaning that the mechanisms of immune priming and progression of human HIT are not replicated.
Radiation-induced thrombocytopenia (RIT)
Despite extensive research in rodents, dogs, and non-human primates, the mechanisms of RIT remain elusive 23. Species and strain-specific effects on radiation sensitivity were reported 22, hindering reliable extrapolation of responses to humans. For instance, the BALB/c mouse strain exhibits a defect in repairing dsDNA, primarily due to a polymorphism in the gene encoding DNA-dependent protein kinase catalytic subunit that plays a critical role in non-homologous end joining. The fact that BALB/c mice are naturally Th2 (ant-inflammatory) biased while C57BL/6 mice tend to have a Th1 (pro-inflammatory) biased immune profile adds to the difficulty of understanding how total body irradiation affects immunity and susceptibility to infections in humans. Additionally, species-specific differences in genetics obscure our understanding of how genetic variants in DNA repair pathways, oxidative stress, inflammatory signaling and tissue remodeling influence the extent and rate of cell loss after irradiation. Identifying the role of these variants would help determine which cancer patients are at risk of developing chronic adverse side effects after treatment 65.
Primary and secondary immune thrombocytopenia (ITP)
Animal models of ITP are not representative of the multitude of etiologies of human ITP (idiopathic, autoimmune, malignancy, viral infection) that may have different mechanisms. In particular, human-relevant immune components and genetic polymorphisms (HLA, Fc receptors) that may explain variable responses to TPO-RA therapy in ITP patients are missing in preclinical in vivo models.
Since anti-platelet antibodies are externally administered in rodent models of ITP, the modeling approach bypasses B cell activation and autoantibody generation, which are central to human ITP pathogenesis, thwarting possibility to understand the mechanisms of this autoimmune disorder.
Contrarily to mice, dogs with spontaneous ITP mirror the idiopathic nature of most human cases. However, the dog model does not fully recapitulate the immunopathogenesis of human ITP. While it modelled the humoral component of platelet destruction, dogs with spontaneous ITP did not show direct evidence of upstream immune dysregulation (Th1/Th2 imbalance, B cell tolerance breakdown, macrophage FcyRIIA/FcyRIIB imbalance) and autoreactive CD8 T cells 67.
The adoptive transfer approach of immunizing CD41-knockout mice with platelets of wild-type mice, and subsequently transferring splenocytes from immunized mice into SCID mice, recapitulated the production of anti-platelet antibodies and T-cell mediated destruction 68. However, this alloimmune model bypasses natural tolerance mechanisms, lacks human-relevant immune regulation and does not enable identification of the antigens targeted in human ITP.
Primary and secondary inherited thrombocytopenia (IT)
In humans, Gray platelet syndrome (GPS) is most commonly inherited in an autosomal recessive pattern 69. The approach of creating Nbeal2-knockout mice reflects the disease etiology since most GPS patients are either homozygous or compound heterozygous for mutations in NBEAL2 gene. Wistar Furth rats 72, on the other hand, do not mirror the genetic cause of human GPS since they lack known a NBEAL2 mutation. Inter-strain differences in Nbeal2-null mice 70, 71 suggest that modifier genes can have an impact on megakaryocyte development and proplatelet formation.
Human MYH9 related disease (MYH9-RD) is an autosomal dominant disorder 73. The fact that genetically engineered heterozygous Myh9 mice do not match the severity of the human MYH9-RD phenotype 74 points to the existence of species-specific compensatory pathways.
ANKRD26-related thrombocytopenia also follows an autosomal dominant pattern of inheritance 12, 53. The inability of mouse models to recapitulate the human disease phenotype can be explained by species-specific differences in genetics. Mouse Ankrd26 gene has a different 5′ UTR sequence and the upstream binding motifs for transcription factors RUNX1 and FLI1 are absent in mice, posing a major barrier to achieving robust construct validity of mouse models of ANKRD26-related thrombocytopenia 54.
Patients with congenital amegakaryocytic thrombocytopenia (CAMT) carry mutations in MPL gene encoding platelet TPO receptor, of which effects range from complete loss (null allele) to partial signaling activity 25, 26. Double knock-out (DKO) Mpl-deficient mice do not model the genetic heterogeneity seen in CAMT patients, only the most severe loss of function end of the spectrum. That even DKO Mpl-deficient mice 75, 76 still fail to faithfully recapitulate the human CAMT phenotype, can be explained by human-specific characteristics of platelet receptors and hematopoietic regulation that are not well understood and are not reproduced in mouse models 36, 40.
Thrombotic thrombocytopenic purpura (TTP)
While many cases of human congenital TTP (cTTP) require a triggering factor to manifest clinically, in contrast to animal models of cTTP, some individuals develop symptoms spontaneously in infancy or early childhood without known external triggers. Over 260 disease-associated ADAMTS13 mutation sites have so far been identified, involving frameshift, missense, and splice site mutations 2, 13, 14. The considerable time and cost required to develop transgenic mouse models prohibits investigation in mice of individual and combined effects of such a high number of mutations. In addition, the species-specific differences described in previous section represent a major barrier for identifying the role of modifier genes (residual ADAMTS13 activity, VWF clearance, inflammatory cytokines, immune tolerance) and understanding how their expression under stress (infection, pregnancy, surgery) affects cTTP development and risk of episodic flares. Inducing acute TTP by injecting Shiga toxin or recombinant human UL‐VWF 77, 78 is not representative of mechanisms that underly TTP induction by human triggering factors. Furthermore, the impact of triggering factors of human cTTP was found to vary across species - in contrast to women with hereditary ADAMTS13 deficiency that get acute TTP on pregnancy, pregnancy did not trigger TTP-like symptoms in ADAMTS13-deficient female mice 81.
A major limitation of all animal models of acquired immune TTP (iTTP), including non-human primate iTTP models, is that they do not reflect true autoimmunity 79, 80. Passive immunization through injection of antibodies against ADAMTS13 bypasses the natural immune sensitization and B-cell activation seen in human iTTP. As a result, these in vivo iTTP models do not replicate loss of immune tolerance, antigen presentation, B cell and plasma cell maturation, and continuous production of anti-ADAMTS13 antibodies. These limitations hinder the ability to understand the mechanisms of autoimmunity in iTTP and develop effective therapies that prevent disease relapse.
Disseminated intravascular coagulation (DIC)
Despite mimicking human triggers (sepsis, trauma, cancer), most animal models of DIC fall short in capturing the complexity and heterogeneity of human-specific coagulation and immune responses. In patients with DIC the disease mechanisms involve a complex cascade of events, including immune activation, endothelial dysfunction, tissue factor expression, and fibrinolysis. In stark contrast, most animal models of DIC tend to be reductionist and mechanistically narrow. For instance, LPS injection and cytokine infusion to mimic sepsis-induced DIC does not replicate pathogen-host interactions that initiate a cascade of events – immune activation, endothelial injury, and dysregulation of coagulation. In the same manner, factor Xa and tissue factor administration are ill suited for studying mechanisms of upstream inflammation, immune signaling, and vascular response. Because they do not reflect the etiological diversity and mechanistic complexity seen in human DIC, animal models of DIC are also inherently limited in their capacity to support the development of personalized therapies 85.
Hemolytic uremic syndrome (HUS)
Several species-specific differences stand in the way of achieving robust construct validity of animal models of HUS. The ability of STEC strains to reach target tissues is influenced by the distribution of Stx-binding Gb3 receptors, the immune status, and the microbiome. Compared to mice, the expression level of Gb3 in renal glomerular endothelium in pigs is more similar to humans. Yet, gnotobiotic piglets infected with human-relevant STEC strain predominantly develop neurological complications rather than severe renal pathology and hematologic features 87, suggesting that the mechanisms underlying HUS in pigs differ from those in humans. Given that are significant inter-species differences in immune system composition and function, model organisms may mount distinct innate and adaptive immune responses to bacterial infection 90. Co-injected with STEC/Stx, LPS has a synergistic effect through immune cell priming, inflammation, vascular permeability, and platelet activation. The fact that HUS modelling in mice routinely requires priming by LPS 86 to develop severe thrombocytopenia, anemia, and renal damage, points to inter-species differences in mechanisms of response to STEC infection. Stx injection in baboons 89 does not capture pathogen-host interactions, which is problematic for understanding the mechanisms of STEC colonization and adherence to target tissues, epithelial signaling, Stx uptake, and immune activation. The gut microbiome in model organisms differs significantly from the microbiome in humans, affecting STEC colonization, immune modulation, and HUS progression in ways that diverge from human HUS pathogenesis 91.
Predictive validity - How well do animal models predict safety and efficacy of therapies in human patients?
Largely owing to decades of reliance on experimental animal models of thrombocytopenia, most available treatments are non-curative and only transiently compensatory. Given high heterogeneity in pathophysiology within thrombocytopenia types, another challenge for developing therapies for thrombocytopenia is the uncertainty in terms of which patients to treat and with which therapy. Drug-induced thrombocytopenia, typical hemolytic uremic syndrome, and disseminated intravascular coagulation are among the most common cases found in intensive care units, and yet effective treatments for these conditions are sorely lacking.
Aplastic anemia (AA)
For patients with idiopathic AA, allogeneic hematopoietic stem cell transplantation (HSCT), with or without adjunct immunosuppressive therapy (IST), is the first-line treatment, especially in children and young adults. HSCT has demonstrated particularly high efficacy in patients with acquired idiopathic AA and hereditary AA 92, 93. The degree of efficacy and safety of HSCT was notably evaluated in chemically-induced (benzene, busulfan) and immune-mediated (injection of allogeneic lymphocytes) animal models of AA (mice, dogs). The risk of graft failure and graft versus host disease remain one of the main challenges in HSCT, necessitating novel strategies to prevent or mitigate these complications. Addition of thrombopoietin receptor agonist (TPO-RA) eltrombopag to IST in adults is considered standard of care for acquired severe aplastic anemia in adults who are ineligible for HSCT 94, but is not recommended for hereditary AA or other AA etiologies. It is also noteworthy that clinical development of AA therapies such as IST agents (antithymocyte globulin, cyclosporine, corticosteroids) did not rely primarily on their success in AA mouse models and were initially investigated in rheumatology and transplantation, underscoring the merit of the human-centric approach based on observations by clinicians 95. Relapse in treated AA patients, which occurs more frequently after IST than HSCT, is another major concern. The fact that animal models of AA do not allow to reveal the specific antigens that trigger and sustain the autoreactive T cell response against HSPC represents a major hindrance for developing antigen-specific therapies.
Drug-induced thrombocytopenia (DIT) - Chemotherapy-induced thrombocytopenia (CIT) - Heparin-induced thrombocytopenia (HIT)
There is no targeted treatment to directly reverse DIT. Typically, the offending drug is immediately discontinued and supportive care is administered through platelet transfusion and intravenous immunoglobulins (IVIG)/corticosteroids 5, 20. In some cases a persistent immune-mediated thrombocytopenia may develop despite drug withdrawal, requiring adapted treatment. A better prediction of DIT for various drugs would allow to identify drug-specific risks early in drug development process or patient-specific risks before exposure to a drug. However, due to above described inter-species differences (immune system, genetic susceptibility), animal testing is a poor predictor of drug-induced immune thrombocytopenia in humans 34, 96.
Chemotherapy-induced thrombocytopenia (CIT) often leads to negative outcomes in cancer patients, and yet, there are no standardized guidelines for preventing or managing CIT. Recombinant TPO reversed thrombocytopenia and increased survival in animal models of chemotherapy and radiation-induced thrombocytopenia 61, 97. However, clinical studies using recombinant TPO were halted following the occurrence of aplastic anemia-like syndrome with development of neutralizing antibodies to TPO in some cancer patients who were treated with PEGylated recombinant human megakaryocyte growth and development factor (PEG-rHuMGDF), a truncated form of TPO 98. The peptibody analogue of TPO, romiplostim, and small molecule drugs eltrombopag, avatrombopag, and lusutrombopag exert their effects by bind to and activating the TPO receptor. Administered following chemotherapy-radiation cycles to mimic human chemotherapy-radiation-induced thrombocytopenia, romistom treatment was effective at accelerating platelet recovery in BDF1 mice 100. However, these TPO-RA have not received approval from the FDA or EMA for treatment of CIT. The National Comprehensive Cancer Network endorsed consideration of romiplostim as one option for patients with CIT 63, 99. The International Society on Thrombosis and Haemostasis (ISTH) Subcommittee on Hemostasis and Malignancy recommends against the use of TPO-RA for the management of CIT in acute myeloid leukemia, HSCT, and lymphoma. In patients receiving chemotherapy for acute myeloid leukemia, there was a trend of worsened survival with eltrombopag compared with placebo. A single phase 3 trial in patients with ovarian, bladder, or lung cancer receiving chemotherapy with severe thrombocytopenia, treatment with avatrombopag did not meet its primary efficacy end point. Similarly, in a phase 2 study of patients with solid tumor diagnosis receiving gemcitabine monotherapy or gemcitabine plus cisplatin/carboplatin, eltrombopag did not lead to a significant improvement in platelet nadir compared with placebo. When considering off-label use of TPO-RA for solid tumors, use of romiplostim over other TPO-RA was preferred 7, 63.
HIT is managed by discontinuing heparin immediately and by administering alternative anticoagulants, such as argatroban, bivalirudin and fondaparinux 1, 9. In transgenic hPF4/FcγRIIA mouse models of HIT, treatment with a selective inhibitor of the tyrosine kinase Syk prevented thrombocytopenia and thrombosis 101. These preclinical results have, however, not translated into approved clinical therapies for HIT in humans. In some cases, PF4 antibodies occur without prior heparin exposure or persist despite heparin discontinuation and treatment with alternative anticoagulants 1, 9. Thrombin inhibitors (argatroban, bivalirudin) do not neutralize the upstream antibody-mediated platelet activation. Immunomodulatory treatments like IVIG and steroids are immunosuppressive but non-specific. The fact that animal models of HIT do not capture the autoimmune mechanisms that generate and sustain antibody production hinders progress toward developing targeted therapies for autoimmune HIT.
Radiation-induced thrombocytopenia (RIT)
Despite decades of preclinical research in rodents, dogs, and non-human primates 23, no FDA or EMA-approved therapies currently exist to prevent or mitigate RIT. Among therapeutic leads that showed promise in animal models of RIT but did not translate to approval for treatment in humans, are 5-androstenediol that had radioprotective effects of in rhesus macaques in total body-irradiation experiments 102. Nor have irradiation experiments in Wistar rats, Sprague-Dawley rats, WAG/RijCmcr rats, Balb/c mice, C3H/He mice, Cdh5-CreERT2 mice, C57BL/6J mice, NWZ rabbits, and rhesus macaques proven effective for developing treatments for radiation-induced tissue fibrosis, a severe adverse effect of radiotherapy that causes irreversible damage to tissues, including the bone marrow stroma 105. Recombinant cytokines to stimulate hematopoietic progenitors, recombinant human TPO to promote megakaryocyte proliferation and TPO-RA (hetrombopag, eltrombopag, romiplostim) to stimulate platelet production were not specifically approved by the FDA or EMA for treating RIT and may be used off-label 103, 104. A better understanding of RIT pathogenesis using human-based systems would help advance discovery of targeted therapies for RIT.
Primary and secondary immune thrombocytopenia (ITP)
Preclinical research and drug discovery for ITP have primarily relied on passive ant-platelet antibody transfer or immunization with platelet antigens in mice (BALB/c, C57BL/6, NOD/SCID, partially humanized transgenic) mice, rabbits, and chimpanzees. IST such as glucocorticoids (prednisone, dexamethasone) and IVIG are the first line treatments for primary and secondary ITP 3. The anti-CD20 monoclonal antibody rituximab that depletes B cells and immunosuppressants vincristine, cyclophosphamide, and cyclosporine A may be used off-label, often in corticosteroid-refractory ITP cases. For reasons covered in the section ‘species-specific differences in platelet receptor’, human and chimpanzee-specific characteristics of the extracellular domain of the thrombopoietin receptor (MPL) had caused major difficulties in developing small-molecule TPO-RA that bind to the human MPL. TPO-RA (eltrombopag, avatrombopag, lusutrombopag, romiplostim), that stimulate platelet production by activating MPL on megakaryocytes, are typically used as second-line or adjunctive therapies in cases where corticosteroids are contraindicated or patients have chronic/relapsing ITP 35. Eltrombopag was approved by the FDA in 2015 for treatment of thrombocytopenia in children one year and older with primary ITP who have had an insufficient response to corticosteroids, immunoglobulins, or splenectomy. Driven by the need for safer alternatives to eltrombopag, which carries a risk of hepatotoxicity, especially in patients with underlying liver disease, FDA's approval of avatrombopag and lusutrombopag for thrombocytopenia in adults with chronic liver disease followed in 2018 106. The same year, the FDA had approved romiplostim treatment by subcutaneous injection in pediatric patients with ITP 107.
Importantly, neither TPO-RA nor IST directly address the underlying mechanism of platelet destruction in ITP, which involves autoantibody-mediated opsonization, Fcγ receptor–mediated phagocytosis, complement activation, CD8 T-cell cytotoxicity, and NK cell dysfunction. TPO-RA increase the platelet output but do not mitigate the autoimmune process and IST reduces autoantibody production without correcting the broader immune dysregulation.
In ITP treatments, IST is effective short term (about 80% of patients respond to treatment) but has notable post-treatment mortality (10 to 15%, primarily due to bleeding) 108 while TPO-RA offer the same rate of platelet response (about 80%) with lower mortality 110. However, both IST and TPO-RA have high, 70-80%, relapse rates 109, 110, underscoring the need for novel disease-modifying therapies.
Furthermore, owing to heterogeneity in underlying etiology (idiopathic, autoimmune, infectious, drug-induced) and patient health status (immune status, bone marrow reserves, comorbid conditions), the efficacy and safety TPO-RA profiles have shown considerable inter-individual variability 111, 112. These multifactorial determinants of treatment response cannot be recapitulated in animal models, but could be incorporated into human-based in vitro systems to predict patient response to specific/combination therapies, guiding personalized therapy selection.
Primary and secondary inherited thrombocytopenia (IT)
Current pharmacologic approaches for IT are non-curative, non-specific, and largely used off-label 12, 113, 114. Individuals with IT may benefit from either short-term or long-term treatment with TPO-RA. Short-term treatment is typically given in preparation for major surgery when the platelet count is below the safe threshold. Phase 2 clinical trials for short-term course of eltrombopag for MYH9-related disease, ANKRD26-related thrombocytopenia, X-linked thrombocytopenia, Wiskott-Aldrich syndrome, monoallelic Bernard-Soulier syndrome, and ITGB3-related thrombocytopenia showed that the majority of patients responded to the drug 113. Differences in the degree of platelet response between the patients with the different forms of IT were observed, with eltrombopag being the most effective in increasing platelet count in MYH9-related disease and monoallelic Bernard-Soulier syndrome.
However, TPO-RA are not curative and hematopoietic stem cell transplantation (HSCT) remains the treatment of choice for achieving long-term increase of platelet count in a number of IT types, particularly in those with bone marrow failure, immunodeficiency, or high risk of malignant transformation - congenital amegakaryocytic thrombocytopenia, Wiskott-Aldrich syndrome, GATA2 deficiency, RUNX1, ETV6, and ANKRD26-related thrombocytopenia. HSCT is not recommended for grey platelet syndrome, since the risks may outweigh the benefits.
Congenital amegakaryocytic thrombocytopenia (CAMT) presents a uniquely difficult therapeutic challenge for patients who are not eligible for HSCT 114. In CAMT eltrombopag is generally ineffective because functional MPL receptors are absent or severely impaired.
The fact that mouse models engineered to carry IT subtype-specific mutations retain functional MPL signaling makes them broadly responsive to TPO-RA, even when their face and construct validity is poor. These models largely fall short in developing and testing mutation-specific therapies (RNA correction, gene editing, small molecule modulators). A notable exception is gene therapy Waskyra for Wiskott-Aldrich syndrome patients with WAS mutation who do not have a suitable HSCT donor, approved by the EMA in 2025 115. Delivered by self-inactivating lentiviral vectors, Waskyra was shown to reduce bleeding, though not all patients achieved full remission. By leveraging patient-derived tissues that carry IT-specific mutations, human-based in vitro platforms offer a powerful means to uncover new therapeutic targets and enable development of personalized therapies for this underserved group of hereditary disorders.
Thrombotic thrombocytopenic purpura (TTP)
Plasma exchange (PEX) is the cornerstone therapy for acquired immune TTP (iTTP), allowing to reduce iTTP-associated mortality from 90% to 20% 1. In PEX, patient plasma containing anti-ADAMTS13 antibodies is removed and replaced by donor plasma containing functional ADAMTS13. Nonetheless, PEX does not target anti-ADAMTS13 antibodies production, and as long-term studies show, PEX alone can be insufficient to prevent persistent or recurrent ADAMTS13 deficiency that may be responsible for 30–40% frequency of relapse after initial PEX treatment 116, 117. Used off-label in iTTP, corticosteroids, like prednisone, and anti‑CD20 monoclonal antibody rituximab are considered as standard adjuncts to PEX in treating acute iTTP. Cohort studies showed that pre-emptive rituximab, administered after PEX, prevented early relapse and increased the time to first relapse, approximately halving relapse rates in the first few years compared to PEX alone 14, 118. However, extended follow-up studies suggest that long-term risk of iTTP relapse remains whether rituximab is given or not 117. For reasons that are not well understood, response to rituximab was found to vary across patients, with race influencing relapse risk and treatment outcomes 119, underscoring the need for personalized treatment strategies for iTTP. Despite apparent mechanistic relevance and promising results in autoimmune mouse models, N-acetylcysteine (reduction of disulfide bonds in ultra-large VWF multimers) and bortezomib (reduction of anti-ADAMTS13 antibody production by plasma cells) did not demonstrate robust evidence of safety and efficacy in Phase 2/3 trials for iTTP 14. In 2019, the FDA had approved caplacizumab - humanized monoclonal antibody fragment that attacks the A1 section of VWF - for the treatment of iTTP in combination with PEX and immunosuppressive therapy 120, 121. While it provides immediate protection against platelet-VWF aggregation, caplacizumab does not address the autoimmune mechanisms in iTTP nor the ADAMTS13 deficiency mechanisms in cTTP. In fact, none of existing therapies for iTTP and cTTP tackle the underlying disease mechanisms, as a consequence of which patients are faced with persistent risk of relapse. This is not surprising knowing that animal models do not recapitulate the underlying mechanisms of human iTTP and cTTP, thwarting possibilities to develop therapies that target root causes of TTP pathogenesis.
Until recently, plasma infusion as supply of functional ADAMTS13, during acute episodes and prophylactically to prevent recurrence, was the standard of care for congenital TTP (cTTP). Gene therapy as potential one-time curative treatment for cTTP, showed protective effect against Shiga toxin-induced TTP in Adamts13 KO mice 122, but were not tested in clinical trials due to safety concerns and manufacturing hurdles. In 2023, the FDA has approved Takeda’s purified recombinant form of ADAMTS13, Adzynma, for both acute treatment and prophylactic use in cTTP 123. Nonetheless, recombinant ADAMTS13 does not correct the genetic mutation responsible for cTTP, and although Adzynma prophylaxis substantially reduces relapse rates compared to plasma infusion, it does not eliminate the need for life-long therapy.
Disseminated intravascular coagulation (DIC)
Since DIC is secondary to conditions like sepsis, trauma, cancer, infections and obstetric complications, managing the trigger - antibiotics for sepsis, surgery for trauma, chemotherapy for cancer - is considered as the first step 16, 124. If platelet count is very low and heavy bleeding is present, patients may receive supportive care such as platelet transfusion and fresh frozen plasma. Despite decades of animal research 15, there is no medication specific for DIC, which is not surprising given the overall poor face and construct validity of animal models of DIC. For instance, in a baboon model of E. Coli-induced sepsis, site-inactivated FVIIa limited lung injury by attenuating coagulation activation 125, but this benefit did not translate into improved outcomes in patients with acute respiratory distress syndrome 126. Similarly, treatment with tissue factor pathway inhibitor improved survival in the same baboon model 127, but had no effect on all-cause mortality in patients with severe sepsis 128. Heparin may be considered in patients with predominant thrombosis but is not routinely used in DIC because of hemorrhage risk. Antifibrinolytics are only used in severe bleeding when fibrinolysis dominates. Eltrompobag does not address the excessive platelet consumption in DIC, and could potentially worsen microvascular thrombi. Human-based models are need to investigate the highly variable pathophysiology of DIC and to develop novel treatments tailored to the clinical phenotype.
Hemolytic uremic syndrome (HUS)
Animal research into HUS has massively failed to translate to specific treatments for HUS patients. To date, Shiga toxin-producing E. coli (STEC) infections remain a significant public health concern. Yet, available treatments for STEC infections mainly focus on supportive care and avoiding medications that may increase the risk of HUS 17, 18. For reasons covered in face and construct validity sections, animal models do not faithfully recapitulate human STEC infection and HUS. As a result, dozens of therapeutic strategies for STEC-induced HUS, including Stx-targeted antibodies, anti-complement component 5 antibodies, and Stx receptor analogues, showed promise in mice, rats, rabbits, piglets, and non-human primates but not in humans 45, 129. In the same manner, of several experimental vaccines that had demonstrated protection against STEC infections in animal models, none were approved for human use 129. Human-based systems are needed to obtain human-relevant insights into STEC-induced HUS pathogenesis and develop effective therapies.
In patients with atypical HUS (aHUS), the complement C5 inhibitor eculizumab reduced morbidity and improved survival 130. Initially developed and approved for paroxysmal nocturnal hemoglobinuria, the indications for eculizumab were subsequently expanded to include aHUS. Importantly, the discovery that complement activation is dysregulated in aHUS came from human genetic, in vitro and clinical studies, not animal models.
Ethical validity - How well do animal experiments align with human ethical principles?
Preclinical
Ethics is a human-specific philosophical concept. Humans assume the right to conduct scientific experiments on animals, despite the fact that animals clearly express non-consent through their behaviour (fleeing, vocalizing, defecating, defense). Animal research is unethical in essence by human standards, since it involves physical constraint, psychological suffering and deprivation of freedom, social interactions, natural environment, and life purpose. In addition to this baseline, experiments inflict severe clinical harm in animals 131:
Table S5: Severity classification of chemical disease models
Thrombotic thrombocytopenic purpura model - Up to severe clinical signs
Table S13: Severity classification of genetically altered (GA) lines
Mortality - GA lines resulting in lethality from 2 weeks post-partum on: Severe
Coagulation defects - GA lines with coagulation defects: Variable degree of severity, depending on expression of clinical signs
GA lines resulting in long-term moderate pain or short-term severe pain: Severe
GA lines resulting in long-term moderate anxiety or short-term severe anxiety: Severe
Table S6: Severity classification of infectious diseases
Viral and bacterial infection: Up to severe clinical signs or long-lasting moderate clinical signs. Sepsis model: Severe
Table S3: Severity classification of surgery and surgical induction of disease
Thermal injury: second/third-degree burn-induced DIC, skin wounds > 10mm body for mice or > 10mm body for rats: Severe
Organ/cell transplantation where rejection/failure may lead to severe distress, death or impairment of the general condition of the animal: Severe
Table S8: Severity classification of other disease models
Irradiation ± bone marrow transplantation - Irradiation with a lethal dose with/without reconstitution of the immune system. Irradiation with a (sub)lethal dose with reconstitution with development of GvHD: Up to severe clinical signs
Clinical
While there is no consensus on whether an unethical act can be justified by a pursuit of a hypothetically ethical outcome, it was suggested that animal research was necessary to advance treatments for human diseases. However, statistics consistently show that clinical success rates of drugs developed and tested in animals is very low, raising the question of whether it is ethical to put the health of patients at risk.
Hematology had a modest 23.9% likelihood of clinical approval over 2011-2020. Hemophilia A and anemia research account for the largest number of transitions within this disease area 132. Phase 1 to 2 transition success rate for hematology disease group was 69.6%, well above the 52% average of all indications, while Phase 2 to 3 transition success rate was 48.1% versus 28.9% average of all indications.
The 3Rs principle rests on the unfounded assumption that animal models possess robust face, construct and predictive validity. It also implies that animal cruelty is acceptable unless alternative methods are applicable. In practice in vitro methods are used alongside, rather than in place of, animal models. Switching to 100% human-based research enables biomedical science without animal suffering, and invites a shift to a conceptually separate framework that excludes animal use entirely.
Intrinsic validity - How well do animal models capture the clinical heterogeneity of the human disease?
For reasons detailed in previous sections, animal models do not recapitulate the clinical heterogeneity and inter-patient variability in thrombocytopenia phenotypes, etiologies, and responses to treatments 1, 2, 3, 4, 5, 6, 8, 9, 10, 12, 14, 15, 18, 23, 26, 54, 63, 66, 67, 82, 85, 95, 112, 114, 117, 129.
Extrinsic validity - How well does animal experimentation generate reliable and reproducible outcomes?
It is often argued that although animal models have severe limitations, animal research enables to gather insights that may be valuable. However, the basic precondition for a hypothetical benefit is not met since the majority of animal experiments is irreproducible 133. Contributing factors include flawed experimental design, variation in animal strains and experimental conditions, and lack of transparency on methodology and results of animal studies.
In a bid to improve the quality of reporting of animal experiments, the ARRIVE - Animal Research: Reporting of In Vivo Experiments - guidelines, were published in 2010 and updated to ARRIVE 2.0 in 2020 134. Nonetheless, and in spite of significant investment in dissemination, various incentives and training of animal researchers, the Arrive guidelines remain poorly implemented 135, 136.
As a result of weak relevance, rigor and reliability of animal studies, erroneous and misleading hypothesis are generated, animal and human lives are needlessly sacrificed and dozens of billions of dollars are annually wasted 137.
Beyond the problem of experimental design and reporting of animal studies, there are deep-seeded cultural reasons that are not likely to be addressed any time soon, such as the "publish or perish" culture that discourages from unbiased analysis and reporting of negative results and rewards exaggerated and overhyped claims 138.
While in vitro methods are not immune to issues of reproducibility, the moral weight of irreproducible animal studies is not the same. Animals are not inanimate, disposable objects - they are sentient beings who suffer. Their value lies not in how they might serve human interests, but in their intrinsic worth as living, feeling individuals.
In Summary
Thrombocytopenia (TP) is a hematologic pathology in which counts of platelets are below the lower limit of normal. Patients with thrombocytopenia are at an increased risk of bleeding, thrombosis, and mortality. It is a clinically heterogenous condition with a multitude of forms, etiologies, and contributing risk factors.
Over the last decades, various TP types were modelled in vivo, causing severe suffering in animals. Yet, largely owing to numerous species-specific features, including in platelet production regulation, receptors, aggregation factors, signaling cascade, genetics, and immunity, animal models of TP do not faithfully recapitulate human TP clinical features and underlying mechanisms.
Despite decades-long biomedical research in animal models of TP, most available treatments are non-curative and only transiently compensatory. Effective therapies for some of the most common conditions found in ICU, such as drug-induced thrombocytopenia, typical hemolytic uremic syndrome, and disseminated intravascular coagulation are critically missing.
Human-based in vitro methods will allow to improve the predictivity of drug-induced adverse effects, deepen the understanding of TP mechanisms, develop effective therapies and tailor solutions to individual TP cases.
How is Human-Based In Vitro the Answer to Advance Biomedical Research into Thrombocytopenia
*To study human-specific features of hematopoiesis and platelet production in healthy and inherited/acquired thrombocytopenia (TP) state, by recapitulating in vitro the human bone marrow niche, using natural/synthetic scaffold to mimic marrow 3D structure, and human adult HSPC, primary marrow niche cells and PBMC/hiPSC-derived HSPC, mesenchymal cells, endothelial cells, osteoblasts, megakaryocytes, and immune cells 139, 140, 141.
*To assess the risk of HSCT-related graft versus host disease (GVHD), by coculturing host immuno-competent gut/liver/skin organoids with donor T cells or by perfusing host gut/liver/skin-on-chip with host antigen-presenting cells and donor T cells. To personalize prophylaxis for GVHD by measuring in vitro the effect of compounds on host tissue injury markers, cytokine release and donor T cells activation.
*To identify autoantigens that trigger the autoreactive T cell response in aplastic anemia (AA), by engineering human/patient-derived bone marrow organoids to express candidate antigens, and employing activation‑induced marker sorting followed by scTCR‑seq, cytotoxicity assay, peptide-MHC tetramer staining, cytokine secretion and other endpoints.
*To investigate the mechanisms of acquired autoimmune attack in AA and test targeted immunomodulation by perfusing patient-derived bone marrow-on-chip with autologous T cells, with and without exposing the cell culture to external triggers (virus, chemical, drug), and measuring T cell activation, tissue apoptosis/fibrosis, and other endpoints.
*To investigate drug-dependent/drug-specific/hapten-dependent-antibody, and other distinct mechanisms of drug-induced immune thrombocytopenia (immune DIT), by modelling drug-platelet-adaptive immunity interactions in a dual human bone marrow-lymph node microfluidic device, that integrates autologous antigen-presenting cells, T cells, B cells, complement and macrophages, and that recapitulates antigen presentation, B‑cell activation, T helper support, class switching, and anti-platelet GP antibody production.
*To predict patient-specific immune DIT and test therapeutic candidates for immune DIT by measuring platelet count, drug-occupancy, macrophage-mediated platelet clearance, and other readouts in a bone marrow-lymph node-on-chip with patient-derived peripheral blood HSPC/megakaryocytes, mononuclear cells, serum/plasma.
*To investigate the mechanisms of non-immune chemotherapy-induced TP (CIT), by leveraging bone marrow-on-chip containing donor HPSC/megakaryocytes, stromal, and endothelial cells, and by measuring megakaryocyte maturation/ploidy, proplatelet formation, platelet count, platelet activation/aggregation, apoptosis, DNA damage, oxidative stress, and other features 143.
*To predict patient-specific CIT arising from inter-individual differences in drug metabolism, by connecting liver-on-chip (with patient-derived iPSC or primary hepatocytes reflecting patient CYP/UGT genotype) to bone marrow-on-chip (with patient-derived peripheral blood CD34 HSPC/iPSC-derived megakaryocytes, serum/plasma), and measuring the downstream effects of liver-generated chemotherapy drug metabolites on bone marrow toxicity, thrombopoiesis and platelet function.
*To test safety and efficacy of prophylactic and rescue drug candidates for CIT, by exposing chemo untreated/treated patient-specific bone marrow-on-chip (patient-derived HSPC, mononuclear cells, serum/plasma) to individual or combined drug candidates, and measuring megakaryopoiesis, platelet count, size and activity, cytotoxicity, platelet-monocytes aggregation, thrombin generation, fibrosis, and other features.
*To predict patient-specific response to heparin, and investigate the mechanism of heparin-induced TP (HIT), by modelling adaptive humoral response to PF4-heparin complexes and platelet activation, in a dual patient-derived bone marrow-lymph-node-on-chip, that integrates autologous antigen-presenting cells, T cells, B cells, complement and macrophages.
*To test safety and efficacy of therapeutic candidates for HIT (especially for cases in which PF4 antibodies occur without prior heparin exposure or persist despite heparin discontinuation) in patient-derived bone marrow-on-chip containing patient platelets and PF4/patient serum containing anti‑PF4 IgG, by measuring platelet activation in the presence of PF4/IgG, thrombin generation, and microthrombus formation in flow channels.
*To model type, dose and duration-dependent radiotherapy-induced thrombocytopenia (RIT), investigate its mechanisms, and test efficacy of radioprotective agents, by measuring platelet production, DNA damage, stromal damage, apoptosis, vascular injury, inflammation in human bone marrow organoids/bone marrow-on-chip (hiPSC-derived HSPC, mesenchymal stromal cells, endothelial cells, osteoblasts, immune cells) 143.
*To predict the risk of radiation-induced chronic adverse side effects by exposing patient-specific bone marrow organoids/bone marrow-on-chip (patient iPSC-derived HSPC, mesenchymal stromal cells, endothelial cells, osteoblasts, immune cells) to radiation at clinically relevant doses 143.
*To identify the genetic variants responsible for increased susceptibility to radiation-induced chronic adverse side effects, by gene editing in human iPSC-derived bone marrow organoids, exposing edited organoids to radiation, and tracking fibrosis, inflammatory, DNA damage and other markers.
*To investigate the mechanisms of idiopathic primary immune thrombocytopenia (ITP) by modelling anti-platelet GP antibody production by B cells and platelet clearance by macrophages in a dual compartment microfluidic device: human bone marrow (containing autologous platelets, macrophages, complement, mesenchymal stromal cells, endothelium)-lymph node (containing autologous B cells, T cells, antigen presenting cells, complement, stromal cells)-on-chip in which platelet antigens (GPIIb/IIIa, GPIb/IX, GPVI, membrane vesicles) are presented on coated stromal surfaces or multivalent beads to stimulate immune cells. To identify the antigens driving the autoimmune attack by selectively displaying GP targets.
*To test safety and efficacy of new therapeutic candidates, combination/sequential therapy and personalized therapy for ITP, by measuring immune cell activation markers, IgG, macrophage phagocytosis, cytotoxicity, platelet count etc. in patient-derived bone marrow-lymph node-on-chip, and correlating in vitro signatures with patient features (bleeding, platelet count, response to drugs, relapse).
*To examine the role of human T helper cells in ITP pathogenesis and identify human-relevant targets that modulate T helper cell-mediated B cell activation, by leveraging single lymph node-on-chip or dual bone marrow-lymph node-on-chip containing PBMC derived from healthy individuals and ITP patients who harbor inborn errors of immunity associated with immune dysregulation (CTLA4 haploinsufficiency, LRBA deficiency, Treg defects etc.)
*To model primary inherited thrombocytopenia (IT), such as Gray platelet syndrome, Wiskott-Aldrich syndrome, Alport syndrome, MYH9-related disease, ANKRD26-related thrombocytopenia, investigate its mechanisms, and develop new therapies, by employing patient peripheral blood/iPSC/bone marrow HSPC-derived megakaryocytes that carry patient-specific mutations and genetic background, and measuring polyploidization, platelet lineage markers, α-granule content, actin/microtubule architecture, proplatelet dynamics, platelet count, signaling pathway responsiveness in vitro 142.
*To model secondary IT, such as Shwachman–Diamond syndrome and congenital amegakaryocytic thrombocytopenia, investigate its mechanisms, and develop new therapies, by engineering patient-derived bone marrow organoids/organ-on-chip that, in addition to patient-derived HSPC carrying patient mutations, contain the bone marrow microenvironment (stroma, endothelium, immune cells) 143.
*To determine the role of individual and combined rare/common variants in IT, through gene editing (SNV, small insertions/deletions, loss/gain-of function) experiments in human HSPC or bone marrow tissues.
*To predict patient response and develop personalized therapies for IT, by recapitulating platelet biogenesis, without and with treatment, in an engineering 3D bone marrow niche containing patient-derived peripheral blood/iPSC-derived megakaryocytes carrying patient-specific mutations 144, 145.
*To recapitulate congenital amegakaryocytic thrombocytopenia patient heterogeneity, define genotype-phenotype relationships, stratify patients and test therapies, by employing HSPC derived from patients with MPL mutations ranging from complete loss of function to partial signaling activity, and measuring JAK2-STAT5 and signaling pathways, MPL localisation in response to TPO, lineage commitment, transcriptional responses to TPO, platelet output and function, and other endpoints 146.
*To model the autoimmune attack against ADAMTS13, platelet sequestration, microthrombi formation and microvascular occlusion in immune thrombotic thrombocytopenic purpura (iTTP) and dissect its mechanisms, by engineering a human endothelium-lymph node-kidney microfluidic device. Endothelial compartment: hiPSC-derived microvascular endothelial cells secreting vWF, plasma-derived ADAMTS13 and platelets. Lymph node compartment: patient plasma PBMCs and stromal cells, enriched with B cells, CD4 T cells, dendritic cells, cytokines, and stimulated by ADAMTS13 antigens presentation. Kidney compartment: glomerular endothelium, podocytes, pericytes.
*To test drug candidates and develop personalized therapies for iTTP by employing patient-derived cells, platelets and plasma, and measuring ADAMTS13 activity, autoantibody titers, cytokines, platelet tethering density, microthrombus volume, occlusion time, barrier damage in response to drugs in an endothelium-lymph node-kidney-on-chip model of iTTP.
*To determine the effect of individual/combined mutations in ADAMTS13 gene on congenital TTP (cTTP) pathogenesis, identify new targets, and test efficacy of therapies, by prime/base editing in hiPSC, differentiating edited hiPSC into liver organoids, perfusing liver organoid effluent to the endothelium compartment of an endothelium-kidney-on-chip device, and measuring ADAMTS13 activity, platelet tethering density, microthrombus volume, occlusion time, barrier damage.
*To recapitulate heterogenous disseminated intravascular coagulation (DIC) pathophysiology and dissect its mechanisms, by modelling its triggers in a dual human endothelium-kidney-on-chip. Trigger zone - endothelium compartment: human microvascular endothelial cells, platelets, plasma, monocytes/macrophages, neutrophils. Triggers: sepsis (LPS, TNF‑α, IL‑6, IL‑1β), trauma (mechanical injury, HMGB1), malignancy (tumor‑derived TF‑rich vesicles, mucins), virus (poly(I:C), interferons), fibrinolysis (mechanical injury, ischemia/perfusion, hemodilution). Injury zone – target organ compartment: endothelial cells and organ-specific cells.
*To identify therapeutic targets, test candidate therapies for DIC and tailor treatments to each pathophysiology type by using patient-derived cells, platelets, plasma, and measuring tissue factor expression, vWF release, platelet aggregation, thrombin generation, inflammation, fibrinolytic markers in response to drugs in an endothelium-kidney-on chip model of DIC.
*To study the epigenetic mechanisms susceptible to regulate immune tolerance, megakaryocytes differentiation, and responses to drugs, infections and other external risk factors in AA, DIT, ITP, TTP, by measuring methylation, chromatin remodeling, chromatin immunoprecipitation, expression of epigenetic regulators, and by analysing the functional effect of epigenetic editing (CRISPR activation/repression, lncRNA) in human HSPC.
*To study the effect of acute/chronic hepatocellular drug-induced liver injury (DILI) on reduction of thrombopoietin (TPO) levels and platelet production in a human bone marrow-liver-on-chip, by exposing the vascular compartment to hepatotoxic drugs, and by measuring biomarkers of hepatocellular damage, TPO secretion, platelet output, and other endpoints.
*To study host-pathogenic E. Coli interactions and investigate the effect of human gut microbiome on E. Coli pathogenicity, by modelling E. Coli infection in human gut/colon organoids or organs-on-chip, with or without live gut microbiome or microbiome metabolites 147.
*To model STEC-caused typical hemolytic uremic syndrome (HUS) and investigate its mechanisms, by recapitulating the effects of STEC infection in a human multiorgan gut-endothelium-kidney-on-chip. Gut compartment: human colonic epithelium, goblet cells. Exposure: live STEC or purified Stx1/Stx2. Endothelial compartment: human microvascular endothelial cells, platelets, monocytes, neutrophils, complement. Kidney compartment: human glomerular microvascular endothelial cells, podocytes.
*To model atypical HUS in a human multiorgan gut-endothelium-kidney-on-chip by simulating causative factors in vitro: patient plasma with complement dysregulation (autoimmune disorders, genetic mutations), tumor-derived microvesicles (malignancies).
*To test novel therapies and personalized treatments for typical and atypical HUS, by assessing responses to drugs in a patient-specific gut-endothelium-kidney-on-chip containing patient-derived colonic epithelium, endothelial cells, podocytes, monocytes and plasma: thrombin generation, fibrin, platelet adhesion, complement deposition, permeability, albumin leak and other endpoints.
Last Updated: October 2025
Even though in vitro methods have inherent limitations, their relevance to human biology far exceeds that of animal research.
Animal model organisms were never comprehensively compared to humans and scientifically validated. As the above examples show, the full extent of inter-species differences is unknown and new species-specific features are discovered every day, often only once the drug has failed to show safety and efficacy in patients. Complementing in vitro methods with animal experiments is not effective for human patients, because species-specific differences prevent reliable integration and translation of results to humans.
While animal research benefits from experimenting on a complete organism, model organisms fail to replicate the interplay of thousands of human-specific features, from molecular level to organism level, and are therefore not representative of the complete human organism.
To address the challenges of individual human-based in vitro models, they can be integrated with other human-based in vitro methods, AI-driven analysis, clinical data, and real-world patient data.
Have you leveraged in vitro methods in unique ways? We would love to hear how! Join the conversation to exchange ideas, collaborate and inspire new directions in human-based science!
References
1. Santoshi RK, Patel R, Patel NS, Bansro V, Chhabra G. A comprehensive review of thrombocytopenia with a spotlight on intensive care patients. Cureus. Aug 2022. https://doi.org/10.7759/cureus.27718
2. Kravitz MS, Shoenfeld Y. Thrombocytopenic conditions—autoimmunity and hypercoagulability: Commonalities and differences in ITP, TTP, HIT, and APS. American Journal of Hematology. 2005; 80(3):232–242. https://doi.org/10.1002/ajh.20408
3. Jinna S, Khandhar PB. Thrombocytopenia. StatPearls - NCBI Bookshelf. Jul 2023. https://www.ncbi.nlm.nih.gov/books/NBK542208/
4. Giudice V, Selleri C. Aplastic anemia: pathophysiology. Seminars in Hematology, Volume 59, Issue 1, Jan 2022, Pages 13-20. https://doi.org/10.1053/j.seminhematol.2021.12.002
5. Bakchoul T, Marini I. Drug-associated thrombocytopenia. Hematology ASH Education Program, Dec 2018, (1): 576–583. https://doi.org/10.1182/asheducation-2018.1.576
6. Vayne C, Guéry E-A, Rollin J, et al. Pathophysiology and Diagnosis of Drug-Induced Immune thrombocytopenia. Journal of Clinical Medicine, Jul 2020, 9(7), 2212. https://doi.org/10.3390/jcm9072212
7. Gao A, Zhang L, Zhong D. Chemotherapy-induced thrombocytopenia: literature review. Discover Oncology. 2023; 14(1). https://doi.org/10.1007/s12672-023-00616-3
8. Zhao X, Shan X, Sui S, Song Q, Cheng M, Zhao Y. Cancer treatment-induced thrombocytopenia: diagnosis, mechanisms and management. Frontiers in Immunology. 2025; 16. https://doi.org/10.3389/fimmu.2025.1634688
9. Arepally GM. Heparin-induced thrombocytopenia. Blood, May 2017, 129 (21): 2864–2872. https://doi.org/10.1182/blood-2016-11-709873
10. Pietras NM, Gupta N, Vaillant AAJ, Pearson-Shaver AL. Immune thrombocytopenia. StatPearls - NCBI Bookshelf. May 2024. https://www.ncbi.nlm.nih.gov/books/NBK562282/
11. Nelson A, Myers K. Shwachman-Diamond Syndrome. GeneReviews® - NCBI Bookshelf. Sep 2024. https://www.ncbi.nlm.nih.gov/books/NBK1756/
12. Noris P, Pecci A. Hereditary thrombocytopenias: a growing list of disorders. Hematology Am Soc Hematol Educ Program, Dec 2017 (1): 385–399. https://doi.org/10.1182/asheducation-2017.1.385
13. Stanley M, Killeen RB, Michalski JM. Thrombotic thrombocytopenic purpura. StatPearls - NCBI Bookshelf. Apr 2023. https://www.ncbi.nlm.nih.gov/books/NBK430721/
14. Joly BS, Coppo P, Veyradier A. Thrombotic thrombocytopenic purpura. Blood, May 2017, 129 (21): 2836–2846. https://doi.org/10.1182/blood-2016-10-709857
15. Popescu NI, Lupu C, Lupu F. Disseminated intravascular coagulation and its immune mechanisms. Blood, Mar 2020, 139 (13): 1973–1986. https://doi.org/10.1182/blood.2020007208
16. Costello RA, Leslie SW, Nehring SM. Disseminated intravascular coagulation. StatPearls - NCBI Bookshelf. May 2024. https://www.ncbi.nlm.nih.gov/books/NBK441834/
17. Bhandari J, Rout P, Sedhai YR. Hemolytic uremic syndrome. StatPearls - NCBI Bookshelf. Oct 2023. https://www.ncbi.nlm.nih.gov/books/NBK556038/
18. Travert B, Rafat C, Mariani P, et al. Shiga Toxin-Associated Hemolytic Uremic Syndrome: specificities of adult patients and implications for critical care management. Toxins, Apr 2021, 13(5), 306. https://doi.org/10.3390/toxins13050306
19. Elghetany MT, Punia JN, Marcogliese AN. Inherited Bone marrow Failure Syndromes: Biology and diagnostic clues. Clinics in Laboratory Medicine, Volume 41, Issue 3, Sep 2021, Pages 417-431. https://doi.org/10.1016/j.cll.2021.04.014
20. Aster RH, Curtis BR, McFarland JG, Bougie DW. Drug-induced immune thrombocytopenia: Pathogenesis, diagnosis, and management. Journal of Thrombosis and Haemostasis, Volume 7, Issue 6, Jun 2009, Pages 911-918. https://doi.org/10.1111/j.1538-7836.2009.03360.x
21. Manus MM, Lamborn K, Khan W, Varghese A, Graef L, Knox S. Radiotherapy-Associated Neutropenia and Thrombocytopenia: Analysis of risk factors and development of a predictive model. Blood, Apr 1997, 89 (7): 2303–2310. https://doi.org/10.1182/blood.V89.7.2303
22. Williams JP, Brown SL, Georges GE, et al. Animal models for medical countermeasures to radiation exposure. Radiation Research, Jan 2010, 173 (4): 557–578. https://doi.org/10.1667/RR1880.1
23. Singh VK, Seed TM. A review of radiation countermeasures focusing on injury-specific medicinals and regulatory approval status: part I. Radiation sub-syndromes, animal models and FDA-approved countermeasures. International Journal of Radiation Biology. 2017; 93(9):851–869. https://doi.org/10.1080/09553002.2017.1332438
24. Qiao J, Li X, Wu Y, Wu X, Zhu F, Liu N, Qi K, Cheng H, Li D, Sun H, et al. An increased expression profile of Th9/IL-9 correlated with Th17/IL-17 in patients with immune thrombocytopenia. Platelets. 2016; 28(3):287–294. https://doi.org/10.1080/09537104.2016.1218454
25. Ballmaier M, Germeshausen M, Schulze H, et al. C-MPL mutations are the cause of congenital amegakaryocytic thrombocytopenia. Blood, Jan 2001, 97 (1): 139–146. https://doi.org/10.1182/blood.V97.1.139
26. Germeshausen M, Ballmaier M. CAMT-MPL: Congenital amegakaryocytic thrombocytopenia caused by MPL mutations - heterogeneity of a monogenic disorder - A comprehensive analysis of 56 patients. Haematologica Volume 106, Issue 9, 1 Sep 2021, Pages 2439-2448. https://doi.org/10.3324/haematol.2020.257972
27. Mestas J, Hughes CCW. Of mice and not men: Differences between mouse and human immunology. The Journal of Immunology, Volume 172, Issue 5, Mar 2004, Pages 2731–2738. https://doi.org/10.4049/jimmunol.172.5.2731
28. Seok J, Warren HS, Cuenca AG, Mindrinos MN, Baker HV, Xu W, Richards DR, McDonald-Smith GP, Gao H, Hennessy L, et al. Genomic responses in mouse models poorly mimic human inflammatory diseases. Proceedings of the National Academy of Sciences. 2013; 110(9):3507–3512. https://doi.org/10.1073/pnas.1222878110
29. Yue F, Cheng Y, Breschi A, Vierstra J, Wu W, Ryba T, Sandstrom R, Ma Z, Davis C, Pope BD, et al. A comparative encyclopedia of DNA elements in the mouse genome. Nature. 2014; 515(7527):355–364. https://doi.org/10.1038/nature13992
30. McKenzie SE, Taylor SM, Malladi P, et al. The role of the human FC receptor FC gamma RIIA in the immune clearance of platelets: a transgenic mouse model. Journal of Immunology, 162(7):4311-8. Apr 1999. https://pubmed.ncbi.nlm.nih.gov/10201963/
31. Junker F, Gordon J, Qureshi O. FC gamma receptors and their role in antigen uptake, presentation, and T cell activation. Frontiers in Immunology. 2020; 11. https://doi.org/10.3389/fimmu.2020.01393
32. Liu X-g, Liu S, Feng Q, Liu X-n, et al. Thrombopoietin receptor agonists shift the balance of FCΓ receptors toward inhibitory receptor IIB on monocytes in ITP. Blood, Aug 2016, 128 (6): 852–861. https://doi.org/10.1182/blood-2016-01-690727
33. Kaifu T, Nakamura A. Polymorphisms of immunoglobulin receptors and the effects on clinical outcome in cancer immunotherapy and other immune diseases: A General review. International Immunology, Volume 29, Issue 7, July 2017, Pages 319–325. https://doi.org/10.1093/intimm/dxx041
34. Clark M, Steger-Hartmann T. A big data approach to the concordance of the toxicity of pharmaceuticals in animals and humans. Regulatory Toxicology and Pharmacology, Volume 96, Jul 2018, Pages 94-105. https://doi.org/10.1016/j.yrtph.2018.04.018
35. Kuter DJ. The biology of thrombopoietin and thrombopoietin receptor agonists. International Journal of Hematology. 2013; 98(1):10–23. https://doi.org/10.1007/s12185-013-1382-0
36. Lee S, Zhan H. Deciphering the differential impact of thrombopoietin/MPL signaling on hematopoietic stem/progenitor cell function in bone marrow and spleen. Stem Cell Reports, Volume 19, Issue 2, P211-223, Feb 2024. https://doi.org/10.1016/j.stemcr.2023.12.004
37. Nakanishi-Matsui M, Zheng Y-W, Sulciner DJ, Weiss EJ, Ludeman MJ, Coughlin SR. PAR3 is a cofactor for PAR4 activation by thrombin. Nature. 2000; 404(6778):609–613. https://doi.org/10.1038/35007085
38. Chackalamannil S. A case history in natural product-based drug discovery: discovery of vorapaxar (ZontivityTM). Medicinal Chemistry Research. 2022; 31(10):1623–1636. https://doi.org/10.1007/s00044-022-02938-3
39. Kuter DJ, Begley CG. Recombinant human thrombopoietin: Basic biology and evaluation of clinical studies. Blood, Nov 2002, 100 (10): 3457–3469. https://doi.org/10.1182/blood.V100.10.3457
40. Erickson-Miller CL, Delorme E, Tian S-S, et al. Preclinical activity of eltrombopag (SB-497115), an oral, nonpeptide thrombopoietin receptor agonist. Stem Cells, Volume 27, Issue 2, Feb 2009, Pages 424–430. https://doi.org/10.1634/stemcells.2008-0366
41. Levy G, Carillo S, Papoular B, et al. MPL mutations in essential thrombocythemia uncover a common path of activation with eltrombopag dependent on W491. Blood, Volume 135, Issue 12, Mar 2020, Pages 948-953. https://doi.org/10.1182/blood.2019003240
42. Balkenhol J, Kaltdorf KV, Mammadova-Bach E, Braun A, Nieswandt B, Dittrich M, Dandekar T. Comparison of the central human and mouse platelet signaling cascade by systems biological analysis. BMC Genomics. 2020; 21(1). https://doi.org/10.1186/s12864-020-07215-4
43. Lenting PJ, Denis CV, Christophe OD. How unique structural adaptations support and coordinate the complex function of Von Willebrand Factor. Blood, Nov 2024, 144 (21): 2174–2184. https://doi.org/10.1182/blood.2023023277
44. Celi AB, Goldstein J, Rosato-Siri MV, Pinto A. Role of globotriaosylceramide in physiology and pathology. Frontiers in Molecular Biosciences. 2022; 9. https://doi.org/10.3389/fmolb.2022.813637
45. Hall G, Kurosawa S, Stearns-Kurosawa DJ. Shiga Toxin Therapeutics: Beyond Neutralization. Toxins, Sep 2017, 9(9), 291. https://doi.org/10.3390/toxins9090291
46. Waterston RH, Lindblad-Toh K, Birney E, Rogers J, Abril JF, Agarwal P, Agarwala R, Ainscough R, Alexandersson M, An P, et al. Initial sequencing and comparative analysis of the mouse genome. Nature. 2002; 420(6915):520–562. https://doi.org/10.1038/nature01262
47. Dolan ME, Baldarelli RM, Bello SM, Ni L, McAndrews MS, Bult CJ, Kadin JA, Richardson JE, Ringwald M, Eppig JT, et al. Orthology for comparative genomics in the mouse genome database. Mammalian Genome. 2015; 26(7–8):305–313. https://doi.org/10.1007/s00335-015-9588-5
48. Salzberg SL. Open questions: How many genes do we have? BMC Biology. 2018; 16(1). https://doi.org/10.1186/s12915-018-0564-x
49. Makałowski W, Zhang J, Boguski MS. Comparative analysis of 1196 orthologous mouse and human full-length mRNA and protein sequences. Genome Research. 1996; 6(9):846–857. https://doi.org/10.1101/gr.6.9.846
50. Yates CM, Sternberg MJE. The effects of Non-Synonymous Single Nucleotide Polymorphisms (NSSNPs) on Protein–Protein interactions. Journal of Molecular Biology, Volume 425, Issue 21, 1 Nov 2013, Pages 3949-3963. https://doi.org/10.1016/j.jmb.2013.07.012
51. Ruffo P, Traynor BJ, Conforti FL. Unveiling the regulatory potential of the non-coding genome: insights from the Human Genome Project to precision Medicine. Genes & Diseases, Volume 12, Issue 6, Nov 2025, 101652. https://doi.org/10.1016/j.gendis.2025.101652
52. Alsheikh AJ, Wollenhaupt S, King EA, Reeb J, Ghosh S, Stolzenburg LR, Tamim S, Lazar J, Davis JW, Jacob HJ. The landscape of GWAS validation; systematic review identifying 309 validated non-coding variants across 130 human diseases. BMC Medical Genomics. 2022; 15(1). https://doi.org/10.1186/s12920-022-01216-w
53. Bluteau D, Balduini A, Balayn N, Currao M, Nurden P, Deswarte C, Leverger G, Noris P, Perrotta S, Solary E, et al. Thrombocytopenia-associated mutations in the ANKRD26 regulatory region induce MAPK hyperactivation. Journal of Clinical Investigation. 2014; 124(2):580–591. https://doi.org/10.1172/jci71861
54. Bera TK, Liu X-F, Yamada M, Gavrilova O, Mezey E, Tessarollo L, Anver M, Hahn Y, Lee B, Pastan I. A model for obesity and gigantism due to disruption of the Ankrd26 gene. Proceedings of the National Academy of Sciences. 2007; 105(1):270–275. https://doi.org/10.1073/pnas.0710978105
55. Huang J, Xu K, Yu L, et al. Immunosuppression characterized by increased treg cell and IL-10 levels in benzene-induced hematopoietic toxicity Mouse model. Toxicology, Volume 464, Dec 2021, 152990. https://doi.org/10.1016/j.tox.2021.152990
56. Mayer A, Podlech J, Kurz S, Steffens HP, Maiberger S, Thalmeier K, Angele P, Dreher L, Reddehase MJ. Bone marrow failure by cytomegalovirus is associated with an in vivo deficiency in the expression of essential stromal hemopoietin genes. Journal of Virology. 1997; 71(6):4589–4598. https://doi.org/10.1128/jvi.71.6.4589-4598.1997
57. Risitano AM. Immune Insights into AA. Blood, May 2017, 129 (21): 2824–2826. https://doi.org/10.1182/blood-2017-03-771493
58. Chen J, Lipovsky K, Ellison FM, Calado RT, Young NS. Bystander destruction of hematopoietic progenitor and stem cells in a mouse model of infusion-induced bone marrow failure. Blood, Sep 2004, 104 (6): 1671–1678. https://doi.org/10.1182/blood-2004-03-1115
59. Ni R, Fan L, Zhang L, Song Y, Wang H, Wang A, Liu B. A mouse model of irradiation and spleen-thymus lymphocyte infusion induced aplastic anemia. Hematology. 2022; 27(1):932–945. https://doi.org/10.1080/16078454.2022.2113356
60. Ulich TR, Del Castillo J, Senaldi G, Hartley C, Molineux G. PEG-RHUMGDF promotes multilineage hematopoietic recovery in myelosuppressed mice. Experimental Hematology, Volume 27, Issue 12, Dec 1999, Pages 1776-1781. https://doi.org/10.1016/S0301-472X(99)00117-4
61. Hokom MM, Lacey D, Kinstler OB, et al. Pegylated megakaryocyte growth and development factor abrogates the lethal thrombocytopenia associated with carboplatin and irradiation in mice. Blood, Dec 1995, 86 (12): 4486–4492. https://doi.org/10.1182/blood.V86.12.4486.bloodjournal86124486
62. Bougie DW, Nayak D, Boylan B, Newman PJ, Aster RH. Drug-dependent clearance of human platelets in the NOD/Scid mouse by antibodies from patients with drug-induced immune thrombocytopenia. Blood, Oct 2010, 116 (16): 3033–3038. https://doi.org/10.1182/blood-2010-03-277764
63. Soff G, Leader A, Al-Samkari H, et al. Management of chemotherapy-induced thrombocytopenia: Guidance from the ISTH Subcommittee on Hemostasis and Malignancy. Journal of Thrombosis and Haemostasis, Volume 22, Issue 1, Jan 2024, Pages 53-60. https://doi.org/10.1016/j.jtha.2023.09.031
64. Reilly MP, Taylor SM, Hartman NK, et al. Heparin-induced thrombocytopenia/thrombosis in a transgenic mouse model requires human platelet factor 4 and platelet activation through FCΓRIIA. Blood, Oct 2001, 98 (8): 2442–2447. https://doi.org/10.1182/blood.V98.8.2442
65. Straub JM, New J, Hamilton CD, Lominska C, Shnayder Y, Thomas SM. Radiation-induced fibrosis: mechanisms and implications for therapy. Journal of Cancer Research and Clinical Oncology. 2015; 141(11):1985–1994. https://doi.org/10.1007/s00432-015-1974-6.
66. Leontyev D, Neschadim A, Branch DR. Cytokine profiles in mouse models of experimental immune thrombocytopenia reveal a lack of inflammation and differences in response to intravenous immunoglobulin depending on the mouse strain. Transfusion. 2014; 54(11):2871–2879. https://doi.org/10.1111/trf.12680
67. LeVine DN, Birkenheuer AJ, Brooks MB, Nordone SK, Bellinger DA, Jones SL, Fischer TH, Oglesbee SE, Frey K, Brinson NS, et al. A novel canine model of immune thrombocytopenia: has immune thrombocytopenia (ITP) gone to the dogs? British Journal of Haematology. 2014; 167(1):110–120. https://doi.org/10.1111/bjh.13005
68. Semple J, Kapur R. The ultimate murine model of immune thrombocytopaenia. Thrombosis and Haemostasis. 2019; 119(03):353–354. https://doi.org/10.1055/s-0039-1678697
69. Nurden AT, Nurden P. The Gray Platelet Syndrome: Clinical spectrum of the disease. Blood Reviews, Volume 21, Issue 1, Jan 2007, Pages 21-36. https://doi.org/10.1016/j.blre.2005.12.003
70. Kahr WHA, Lo RW, Li L, et al. Abnormal megakaryocyte development and platelet function in NBEal2−/− mice. Blood, Nov 2013, 122 (19): 3349–3358. https://doi.org/10.1182/blood-2013-04-499491
71. Deppermann C, Cherpokova D, Nurden P, Schulz J-N, Thielmann I, Kraft P, Vögtle T, Kleinschnitz C, Dütting S, Krohne G, et al. Gray platelet syndrome and defective thrombo-inflammation in Nbeal2-deficient mice. Journal of Clinical Investigation. 2013; 123(8):3331–3342. https://doi.org/10.1172/jci69210
72. Jackson CW, Hutson NK, Steward SA, Saito N, Cramer EM. Platelets of the Wistar Furth rat have reduced levels of alpha-granule proteins. An animal model resembling gray platelet syndrome. Journal of Clinical Investigation. 1991; 87(6):1985–1991. https://doi.org/10.1172/jci115226
73. Pecci A, Ma X, Savoia A, Adelstein RS. MYH9: Structure, functions and role of non-muscle myosin IIA in human disease. Gene Volume, 664, Jul 2018, Pages 152-167. https://doi.org/10.1016/j.gene.2018.04.048
74. Mhatre AN, Li Y, Bhatia N, Wang KH, Atkin G, Lalwani AK. Generation and Characterization of Mice with Myh9 Deficiency. NeuroMolecular Medicine. 2007; 9(3):205–215. https://doi.org/10.1007/s12017-007-8008-8
75. Alexander WS, Roberts AW, Nicola NA, Li R, Metcalf D. Deficiencies in progenitor cells of multiple hematopoietic lineages and defective megakaryocytopoiesis in mice lacking the thrombopoietic receptor C-MPL. Blood, Mar 1996, 87 (6): 2162–2170. https://doi.org/10.1182/blood.V87.6.2162.bloodjournal8762162
76. Carver-Moore K, Broxmeyer HE, Luoh SM, Cooper S, Peng J, Burstein SA, Moore MW, De Sauvage FJ. Low levels of erythroid and myeloid progenitors in thrombopoietin-and C- MPL-deficient mice. Blood, Aug 1996, 88 (3): 803–808. https://doi.org/10.1182/blood.V88.3.803.803
77. Motto DG, Chauhan AK, Zhu G, Homeister J, Lamb CB, Desch KC, Zhang W, Tsai HM, Wagner DD, Ginsburg D. Shigatoxin triggers thrombotic thrombocytopenic purpura in genetically susceptible ADAMTS13-deficient mice. Journal of Clinical Investigation. 2005; 115(10):2752–2761. https://doi.org/10.1172/jci26007
78. Schiviz A, Wuersch K, Piskernik C, et al. A new mouse model mimicking thrombotic thrombocytopenic purpura: Correction of symptoms by Recombinant Human ADAMTS13. Blood, Jun 2012, 119 (25): 6128–6135. https://doi.org/10.1182/blood-2011-09-380535
79. Deforche L, Tersteeg C, Roose E, Vandenbulcke A, Vandeputte N, Pareyn I, De Cock E, Rottensteiner H, Deckmyn H, De Meyer SF, et al. Generation of Anti-Murine ADAMTS13 antibodies and their application in a mouse model for acquired thrombotic thrombocytopenic purpura. PLoS ONE. 2016; 11(8):e0160388. https://doi.org/10.1371/journal.pone.0160388
80. Feys HB, Roodt J, Vandeputte N, et al. Thrombotic thrombocytopenic purpura directly linked with ADAMTS13 inhibition in the baboon (Papio ursinus). Blood, Sep 2010, 116 (12): 2005–2010. https://doi.org/10.1182/blood-2010-04-280479
81. Banno F, Kokame K, Okuda T, et al. Complete deficiency in ADAMTS13 is prothrombotic, but it alone is not sufficient to cause thrombotic thrombocytopenic purpura. Blood, Apr 2006, 107 (8): 3161–3166. https://doi.org/10.1182/blood-2005-07-2765
82. Berthelsen LO, Kristensen AT, Tranholm M. Animal models of DIC and their Relevance to Human DIC: a Systematic review. Thrombosis Research, Volume 128, Issue 2, Aug 2011, Pages 103-116. https://doi.org/10.1016/j.thromres.2010.12.002
83. Giles AR, Nesheim ME, Mann KG. Studies of Factors V and VIII:C in an animal model of disseminated intravascular coagulation. Journal of Clinical Investigation. 1984; 74(6):2219–2225. https://doi.org/10.1172/jci111648
84. Suga Y, Kubo A, Katsura H, Staub Y, Tashiro K, Yamada S, Morishita E, Asakura H. Detailed exploration of pathophysiology involving inflammatory status and bleeding symptoms between lipopolysaccharide- and tissue factor-induced disseminated intravascular coagulation in rats. International Journal of Hematology. 2021; 114(2):172–178. https://doi.org/10.1007/s12185-021-03158-y
85. Toh CH, Alhamdi Y. Current consideration and management of disseminated intravascular coagulation. Hematology ASH Educ. Program, Dec 2013, (1): 286–291. https://doi.org/10.1182/asheducation-2013.1.286
86. Karpman D, Connell H, Svensson M, Scheutz F, Aim P, Svanborg C. The role of lipopolysaccharide and shiga-like toxin in a mouse model of Escherichia coli O157:H7 infection. The Journal of Infectious Diseases, Volume 175, Issue 3, Mar 1997, Pages 611–620. https://doi.org/10.1093/infdis/175.3.611
87. Donohue-Rolfe A, Kondova I, Mukherjee J, Chios K, Hutto D, Tzipori S. Antibody-Based Protection of Gnotobiotic Piglets Infected with Escherichia coli O157:H7 against Systemic Complications Associated with Shiga Toxin 2. Infection and Immunity. 1999; 67(7):3645–3648. https://doi.org/10.1128/iai.67.7.3645-3648.1999
88. Kang G, Pulimood AB, Koshi R, et al. A monkey model for enterohemorrhagic Escherichia coli infection. The Journal of Infectious Diseases, Volume 184, Issue 2, Jul 2001, Pages 206–210. https://doi.org/10.1086/322011
89. Taylor FB, Tesh VL, DeBault L, et al. Characterization of the baboon responses to Shiga-Like toxin: Descriptive study of a new primate model of toxic responses to STX-1. The American Journal of Pathology, Volume 154, Issue 4, Apr 1999, Pages 1285-1299. https://doi.org/10.1016/S0002-9440(10)65380-1
90. Cheng Y, Song L, Huang Y, Xiong Y, Zhang X, Sun H, Zhu X, Meng G, Xu J, Ren Z. Effect of enterohaemorrhagic Escherichia coli O157:H7‐specific enterohaemolysin on interleukin‐1β production differs between human and mouse macrophages due to the different sensitivity of NLRP3 activation. Immunology. 2015;145(2):258–267. https://doi.org/10.1111/imm.12442
91. Tovaglieri A, Sontheimer-Phelps A, Geirnaert A, et al. Species-specific enhancement of enterohemorrhagic E. coli pathogenesis mediated by microbiome metabolites. Microbiome, Mar 2019, Volume 7, Article Number 43. https://doi.org/10.1186/s40168-019-0650-5
92. Piekarska A, Pawelec K, Szmigielska-Kapłon A, Ussowicz M. The state of the art in the treatment of severe aplastic anemia: immunotherapy and hematopoietic cell transplantation in children and adults. Frontiers in Immunology. 2024; 15:1378432. https://doi.org/10.3389/fimmu.2024.1378432
93. Lum SH, Eikema D-J, Piepenbroek B, et al. Outcomes of hematopoietic stem cell transplantation in 813 pediatric patients with fanconi anemia. Blood, Sep 2024, 144 (12): 1329–1342. https://doi.org/10.1182/blood.2023022751
94. Yang B, Wang L, Fu L, Chen M, Ma J, Han B. Adding eltrombopag to intensive immunosuppressive therapy for severe aplastic anaemia may help adult patients achieve outcomes similar to paediatric patients. Leukemia. 2024; 39(1):261–264. https://doi.org/10.1038/s41375-024-02450-0
95. Scheinberg P, Chen J. Aplastic anemia: What have we learned from animal models and from the clinic. Seminars in Hematology, Apr 2013, 50(2):156-164. https://doi.org/10.1053/j.seminhematol.2013.03.028
96. Downing NS, Shah ND, Aminawung JA, et al. Postmarket safety events among novel therapeutics approved by the US Food and Drug Administration between 2001 and 2010. JAMA, May 2017, 317;(18):1854-1863. https://doi.org/10.1001/jama.2017.5150
97. Mouthon M-A, Van Der Meeren A, Vandamme M, Squiban C, Gaugler M-H. Thrombopoietin protects mice from mortality and myelosuppression following high-dose irradiation: importance of time scheduling. Canadian Journal of Physiology and Pharmacology. 2002; 80(7):717–721. https://doi.org/10.1139/y02-090
98. Basser RL, O’Flaherty E, Green M, et al. Development of pancytopenia with neutralizing antibodies to thrombopoietin after multicycle chemotherapy supported by megakaryocyte growth and development factor. Blood, Apr 2002, 99 (7): 2599–2602. https://doi.org/10.1182/blood.V99.7.2599
99. Al-Samkari H. Thrombopoietin receptor agonists for chemotherapy-induced thrombocytopenia: A new solution for an old problem. Hematology ASH Education Program, Dec 2022, (1): 286–295. https://doi.org/10.1182/hematology.2022000374
100. McElroy PL, Wei P, Buck K, et al. Romiplostim promotes platelet recovery in a mouse model of multicycle chemotherapy-induced thrombocytopenia. Experimental Hematology, Volume 43, Issue 6, Jun 2015, Pages 479-487. https://doi.org/10.1016/j.exphem.2015.02.004
101. Reilly MP, Sinha U, André P, et al. PRT-060318, a novel SYK inhibitor, prevents heparin-induced thrombocytopenia and thrombosis in a transgenic mouse model. Blood, Feb 2011, 117 (7): 2241–2246. https://doi.org/10.1182/blood-2010-03-274969
102. Stickney DR, Dowding C, Authier S, Garsd A, Onizuka-Handa N, Reading C, Frincke JM. 5-Androstenediol improves survival in clinically unsupported rhesus monkeys with radiation-induced myelosuppression. International Immunopharmacology, Volume 7, Issue 4, Apr 2007, Pages 500-505. https://doi.org/10.1016/j.intimp.2006.12.005
103. Yan D, Yang J, Gao Y, et al. Combination of thrombopoietin receptor agonist and recombinant human thrombopoietin for treating cancer therapy induced thrombopenia. Blood, Nov 2022, 140 (Supplement 1): 8420–8421. https://doi.org/10.1182/blood-2022-164490
104. Li G, Huang A, Yang Q, Chen Y, Jiang Y, Liu Z. Effectiveness and safety of hetrombopag in the management of Radiotherapy-Induced thrombocytopenia in patients with gynecological malignancies. Blood, Nov 2023, 142 (Supplement 1): 5431. https://doi.org/10.1182/blood-2023-182634
105. Yu Z, Xu C, Song B, Zhang S, Chen C, Li C, Zhang S. Tissue fibrosis induced by radiotherapy: current understanding of the molecular mechanisms, diagnosis and therapeutic advances. Journal of Translational Medicine. 2023; 21(1):708. https://doi.org/10.1186/s12967-023-04554-0
106. FDA, U.S. Food and Drug Administration. Research C for DEA. FDA approves avatrombopag for thrombocytopenia in adults with chronic liver disease. May 2018. https://www.fda.gov/drugs/resources-information-approved-drugs/fda-approves-avatrombopag-thrombocytopenia-adults-chronic-liver-disease
107. FDA, U.S. Food and Drug Administration. Research C for DEA. FDA approves romiplostim for pediatric patients with immune thrombocytopenia. Dec 2018. https://www.fda.gov/drugs/fda-approves-romiplostim-pediatric-patients-immune-thrombocytopenia
108. Yan X, Yun Z, Tian L, Li H, Tian S, Wang J. Immune thrombocytopenia: a review of pathogenesis and current treatment. Discover Medicine. 2024;1(1). https://doi.org/10.1007/s44337-024-00040-8
109. Wang X, Liu X, Ma Y, et al. Predictive Factors of relapse for primary immune thrombocytopenia in Adults: A Single-Center Retrospective Cohort Analysis. Blood, Nov 2022, 140 (Supplement 1): 2214–2215. https://doi.org/10.1182/blood-2022-164457
110. Luo L, Jin S, Song Z, Chong G, Ding H, Zeng S, Dai M, Wu M. Bridging the gaps between randomized controlled trials and real-world use of thrombopoietin receptor agonists for adult primary immune thrombocytopenia: a systematic review and meta-analysis. Frontiers in Medicine. 2025; 12:1667457. https://doi.org/10.3389/fmed.2025.1667457
111. Soulard M, Galicier L, Mahlaoui N, et al. Efficacy and safety of TPO receptor agonists in treatment of ITP associated with predominantly antibody deficiencies. Blood Advances, Dec 2024, 8 (24): 6171–6182. https://doi.org/10.1182/bloodadvances.2024014370
112. Liu Y, Zhang N, Jiang T, Li Y, Xiong Y, Liu Y. Efficacy and safety and analysis of thrombopoietin receptor agonists for the treatment of immune thrombocytopenia in adults: analysis of a systematic review and network meta-analysis of randomized controlled trials and results of real-world safety data. Frontiers in Medicine. 2025; 12:1531824. https://doi.org/10.3389/fmed.2025.1531824
113. Zaninetti C, Gresele P, Bertomoro A, Klersy C, De Candia E, Veneri D, Barozzi S, Fierro T, Alberelli MA, Musella V, et al. Eltrombopag for the treatment of inherited thrombocytopenias: a phase II clinical trial. Haematologica. 2019; 105(3):820–828. https://doi.org/10.3324/haematol.2019.223966
114. Balduini CL. Treatment of inherited thrombocytopenias. Haematologica, Vol. 107 No. 6, Jun 2022. 2022:1278–1292. https://doi.org/10.3324/haematol.2022.280856
115. European Medicines Agency (EMA). First gene therapy to treat rare disease Wiskott-Aldrich syndrome. Nov 2025. https://www.ema.europa.eu/en/news/first-gene-therapy-treat-rare-disease-wiskott-aldrich-syndrome
116. Mannucci PM, Bohm M, Scharrer I, Scheiflinger F. Patterns of changes of anti‐ADAMTS13 after plasma exchange. Journal of Thrombosis and Haemostasis, Volume 4, Issue 6, 1405 - 1406. https://doi.org/10.1111/j.1538-7836.2006.01960.x
117. Doyle AJ, Stubbs MJ, Dutt T, et al. Long-term risk of relapse in immune-mediated thrombotic thrombocytopenic purpura and the role of anti-CD20 therapy. Blood, Jan 2023, 141 (3): 285–294. https://doi.org/10.1182/blood.2022017023
118. Zwicker JI, et al. Adjuvant low-dose rituximab and plasma exchange for acquired TTP. Blood, Sep 2019, 134 (13): 1106–1109. https://doi.org/10.1182/blood.2019000795
119. Chaturvedi S, Antun AG, Farland AM, et al. Race, Rituximab, and relapse in TTP. Blood, Sep 2022, 140 (12): 1335–1344. https://doi.org/10.1182/blood.2022016640
120. FDA U.S. Food And Drug Administration. FDA approved caplacizumab-yhdp. Mar 2019. https://www.fda.gov/drugs/resources-information-approved-drugs/fda-approved-caplacizumab-yhdp
121. Pishko AP, Li A, Cuker A. Immune thrombotic thrombocytopenic purpura. JAMA, May 2025; 334;(6):517-529. https://doi.org/10.1001/jama.2025.3807
122. Jin S-Y, Xiao J, Bao J, Zhou S, Wright JF, Zheng XL. AAV-mediated expression of an ADAMTS13 variant prevents shigatoxin-induced thrombotic thrombocytopenic purpura. Blood, May 2013, 121 (19): 3825–3829. https://doi.org/10.1182/blood-2013-02-486779
123. FDA U.S. Food And Drug Administration Office of the Commissioner. FDA Approves First Treatment for Patients with Rare Inherited Blood Clotting Disorder. Nov 2023. https://www.fda.gov/news-events/press-announcements/fda-approves-first-treatment-patients-rare-inherited-blood-clotting-disorder
124. Pène F, Russell L, Aubron C. Thrombocytopenia in the intensive care unit: diagnosis and management. Annals of Intensive Care. 2025; 15(1):25. https://doi.org/10.1186/s13613-025-01447-x
125. Welty-Wolf KE, Carraway MS, Miller DL, Ortel TL, Ezban M, Ghio AJ, Idell S, Piantadosi CA. Coagulation blockade prevents sepsis-induced respiratory and renal failure in baboons. American Journal of Respiratory and Critical Care Medicine. 2001; 164(10):1988–1996. https://doi.org/10.1164/ajrccm.164.10.2105027
126. Vincent J-L, Artigas a, Petersen LC, Meyer C. A multicenter, randomized, double-blind, placebo-controlled, dose-escalation trial assessing safety and efficacy of active site inactivated Recombinant Factor VIIA in subjects with acute lung injury or acute respiratory distress syndrome. Critical Care Medicine, 37(6):P 1874-1880, Jun 2009. https://doi.org/10.1097/CCM.0b013e31819fff2c
127. Creasey AA, Chang AC, Feigen L, Wün TC, Taylor FB, Hinshaw LB. Tissue factor pathway inhibitor reduces mortality from Escherichia coli septic shock. Journal of Clinical Investigation. 1993; 91(6):2850–2860. https://doi.org/10.1172/jci116529
128. Abraham E, Reinhart K, Opal S. Efficacy and Safety of Tifacogin (Recombinant Tissue Factor Pathway Inhibitor) in Severe Sepsis A Randomized Controlled Trial. JAMA, Jul 2003, 290;(2):238-247. https://doi.org/10.1001/jama.290.2.238
129. Mühlen S, Dersch P. Treatment Strategies for Infections With Shiga Toxin-Producing Escherichia coli. Frontiers in Cellular and Infection Microbiology. 2020; 10:169. https://doi.org/10.3389/fcimb.2020.00169
130. Legendre CM, Licht C, Muus P, Greenbaum LA, Babu S, Bedrosian C, Bingham C, Cohen DJ, Delmas Y, Douglas K, et al. Terminal complement inhibitor eculizumab in atypical Hemolytic–Uremic syndrome. New England Journal of Medicine. 2013; 368(23):2169–2181. https://doi.org/10.1056/nejmoa1208981
131. De Vleeschauwer S , Lambaerts K, Hernot S, Karlijn Debusschere K. Severity classification of laboratory animal procedures in two Belgian academic institutions. Animals, 13(16), 2581, Aug 2023. https://doi.org/10.3390/ani13162581
132. BIO. Clinical development success rates and contributing Factors 2011-2020 | https://www.bio.org/clinical-development-success-rates-and-contributing-factors-2011-2020
133. Freedman LP, Cockburn IM, Simcoe TS. The Economics of Reproducibility in Preclinical research. PLoS Biology. 2015; 13(6):e1002165. https://doi.org/10.1371/journal.pbio.1002165
134. ARRIVE Guidelines. https://arriveguidelines.org/
135. Du Sert NP, Hurst V, Ahluwalia A, Alam S, Avey MT, Baker M, Browne WJ, Clark A, Cuthill IC, Dirnagl U, et al. The ARRIVE guidelines 2.0: Updated guidelines for reporting animal research. BMC Veterinary Research. 2020; 16(1). https://doi.org/10.1186/s12917-020-02451-y
136. Mazoit JX. Statistics in Biology: A survey of the three major multidisciplinary journals. BioRxiv, Feb 2025. https://doi.org/10.1101/2025.02.04.636422
137. Yarborough M, Bredenoord A, D’Abramo F, Joyce NC, Kimmelman J, Ogbogu U, Sena E, Strech D, Dirnagl U. The bench is closer to the bedside than we think: Uncovering the ethical ties between preclinical researchers in translational neuroscience and patients in clinical trials. PLoS Biology. 2018; 16(6):e2006343. https://doi.org/10.1371/journal.pbio.2006343
138. Smaldino PE, McElreath R. The natural selection of bad science. Royal Society Open Science. 2016;3(9):160384. https://doi.org/10.1098/rsos.160384
139. Di Buduo CA, Wray LS, Tozzi L, et al. Programmable 3D silk bone marrow niche for platelet generation ex vivo and modeling of megakaryopoiesis pathologies. Blood, Apr 2015, 125 (14): 2254–2264. https://doi.org/10.1182/blood-2014-08-595561
140. Roy A, Wang G, Iskander D, et al. Transitions in lineage specification and gene regulatory networks in hematopoietic stem/progenitor cells over human development. Cell Reports, Volume 36, Issue 11, 109698, Sep 2021. https://doi.org/10.1016/j.celrep.2021.109698
141. Glaser DE, Curtis MB, Sariano PA, et al. Organ-on-a-chip model of vascularized human bone marrow niches. Biomaterials, Volume 280, Jan 2022, 121245. https://doi.org/10.1016/j.biomaterials.2021.121245
142. Di Buduo CA, Alberelli MA, Glembotsky AC, Podda G, Lev PR, Cattaneo M, Landolfi R, Heller PG, Balduini A, De Candia E. Abnormal proplatelet formation and emperipolesis in cultured human megakaryocytes from gray platelet syndrome patients. Scientific Reports. 2016; 6(1). https://doi.org/10.1038/srep23213
143. Chou DB, Frismantas V, Milton Y, David R, Pop-Damkov P, Ferguson D, MacDonald A, Bölükbaşı ÖV, Joyce CE, Teixeira LSM, et al. On-chip recapitulation of clinical bone marrow toxicities and patient-specific pathophysiology. Nature Biomedical Engineering. 2020; 4(4):394–406. https://doi.org/10.1038/s41551-019-0495-z
144. Di Buduo CA, Laurent P-A, Zaninetti C, et al. Miniaturized 3D bone marrow tissue model to assess response to thrombopoietin-receptor agonists in patients. eLife, Jun 2021. https://doi.org/10.7554/eLife.58775
145. Di Buduo CA, Lunghi M, Kuzmenko V, Laurent P, Della Rosa G, Del Fante C, Nogare DED, Jug F, Perotti C, Eto K, et al. Bioprinting Soft 3D Models of Hematopoiesis using Natural Silk Fibroin‐Based Bioink Efficiently Supports Platelet Differentiation. Advanced Science. 2024; 11(18):e2308276. https://doi.org/10.1002/advs.202308276
146. Hirata S, Takayama N, Jono-Ohnishi R, Endo H, Nakamura S, Dohda T, Nishi M, Hamazaki Y, Ishii E-I, Kaneko S, et al. Congenital amegakaryocytic thrombocytopenia iPS cells exhibit defective MPL-mediated signaling. Journal of Clinical Investigation. 2013; 123(9):3802–3814. https://doi.org/10.1172/jci64721
147. Karve SS, Pradhan S, Ward DV, Weiss AA. Intestinal organoids model human responses to infection by commensal and Shiga toxin producing Escherichia coli. PLoS ONE. 2017; 12(6):e0178966. https://doi.org/10.1371/journal.pone.0178966