Keywords
blood platelet disorders - high-throughput nucleotide sequencing - platelets
Schlüsselwörter
angeborene Thrombozytenstörung - Hochdurchsatzsequenzierung - Thrombozyten - Blutplättchen
Introduction
Platelets are main players in hemostasis. They do not only act as sentinels of the
vascular integrity but can also seal wounds by forming a hemostatic plug, a process
referred to as primary hemostasis.[1] In addition, platelets can build a procoagulant surface, where certain steps of
the coagulation cascade take place thus bridging the primary with the secondary hemostasis.[2] Ectopic platelet activation, in contrast, can result in thrombotic events that eventually
lead to myocardial infarction or stroke. The equilibrium of hemostasis and thrombosis
has been the center of platelet research for decades, but it has been increasingly
recognized that platelets also play an important role as immune mediators, where tissue-
and context-specific roles become more clear, as indicated by thrombo-inflammation
or immunothrombosis (like in COVID-19).[3]
[4] Platelets have an overall short half-life of 8 to 10 days before they become sequestered
by the reticuloendothelial system. They are thus continuously replenished from megakaryocytes
(MKs), the immediate precursor cell in the bone marrow, with ∼1011 generated per day.[5] Defects in platelet count or function are typically referred to as inherited platelet
disorders (IPDs). They can be divided into (1) inherited thrombocytopenias, in which
the paucity of platelets is a consequence of insufficient production (either amegakaryocytic
or hypomegakaryocytic) or a shortened lifespan, and (2) inherited platelet function
defects (PFDs) that are typically associated with an overall normal platelet count.
The underlying causes of PFDs are dysfunctional surface receptors, defects in the
signaling cascades, cytoskeletal alterations, inadequate number of internal granules,
their contents, or an inappropriate release. Defects in MK-specific transcription
factors can affect multiple of those mentioned aspects. IPDs can occur as an isolated
feature or as a syndromic disorder with additional symptoms. Some IPDs have specific
pathognomonic features such as Glanzmann's thrombasthenia (GT), Bernard-Soulier syndrome
(BSS), or Congenital Amegakaryocytic Thrombocytopenia (CAMT) of which the relevant
gene (or genes) has been well studied.[6]
[7]
[8]
[9] Other examples for IPDs that have been deciphered in the last two decades include
Gray platelet syndrome[10] or thrombocytopenia-absent radius syndrome.[11]
[12] So far approximately 60 to 80 genes are considered to be involved in IPDs.[13] However, the genetic and experimental evidence for many of them may still be considered
to be incomplete. International consortia are thus constantly evaluating the underlying
evidence, which is finally reflected by a list of GoldVariants that are the “Tier
1” level of genes to be taken into account. New genes are considered, but genes are
also removed when the evidence is insufficient.[5] The relevance of clinical complications in patients with IPDs is highly variable,
even within the same disease type, ranging from almost negligible to life threatening.
As some patients experience massive bleeding in response to surgery or trauma, it
is important to know the underlying condition.[14] Consequently, an early and accurate diagnosis of the disorder and a close medical
follow-up of the affected patient is of great relevance. Although research over the
last three decades has led to the discovery of a relevant number of genes harboring
variants responsible for IPDs,[13] evidence of pathogenicity of some of these genes remains limited. More importantly,
more than a third of affected patients still do not receive a sound molecular (genetic)
diagnosis.[15] In this article, we aim to elucidate the state-of-the-art high-throughput sequencing
(HTS) in the diagnosis of IPDs, discuss advantages and disadvantages of this method,
and describe challenges that have to be considered by those involved in the diagnosis
and management of patients with IPDs.
Traditional Diagnosis of IPDs and Its Limitations
Traditional Diagnosis of IPDs and Its Limitations
Several guidelines and expert committees recommend the following evaluations as first-line
diagnostics of IPDs: (1) clinical investigation of the personal and familial bleeding
history with an emphasis on unexplained or extensive bleeding events; (2) physical
examination assessing for typical bleeding manifestations and for potential syndromic
features (e.g., hearing loss, cardiac anomalies, facial dysmorphia or skeletal abnormalities,
ocular involvement, skin discoloration, and mental retardation); (3) comprehensive
laboratory testing for defects in secondary hemostasis; (4) exclusion of von Willebrand
disease; and (5) evaluation of a complete blood cell count (CBC) and blood smear,
especially focusing on platelet count, size (mean platelet volume [MPV]), and morphology
([Fig. 1]).[16]
[17]
[18] In general, phenotypic and functional analyses usually start with relatively simple,
widespread, and largely nonspecific tests (e.g., CBC, blood smear, prothrombin time,
von Willebrand factor). Subsequently, more specific and complex methods (e.g., platelet
aggregometry, flow cytometry analysis, immunofluorescence staining) are typically
used.[19]
[20]
[21]
[22] However, the requirement of relatively large volumes of freshly drawn blood (especially
for platelet function analyses) and the availability of these tests in a small number
of highly specialized laboratories limit the speedy and sound diagnosis of IPDs, although
much progress has been made regarding the standards and quality control of diagnostics
in pediatric patients.[16]
[20]
[23]
Fig. 1 Diagnostic algorithm for inherited platelet disorders (IPDs). IPF, immature platelet
fraction; MPV, mean platelet volume; Plt, platelets; WES, whole exome sequencing;
WGS, whole genome sequencing.
Indications for Genetic Testing
Indications for Genetic Testing
A worldwide laboratory survey of the International Society on Thrombosis and Haemostasis (ISTH) revealed that even if a broad range of analytical methods is applied, the
exact underlying defect cannot be identified in more than one-third of patients with
confirmed abnormalities of platelet function.[24] However, a low platelet count does not exclude an associated altered platelet function;
even if the functional defect is minor, it could contribute to the bleeding phenotype.[25] Indeed, a definite diagnosis of IPDs can only be achieved by identifying the underlying
molecular pathology and thus, genetic testing might be introduced much earlier in
the diagnostic process ([Fig. 1]). This accounts particularly for patients with a bleeding tendency that is already
indicative of a certain platelet dysfunction (e.g., GT), a positive family history,
or conspicuous diagnostics in the functional testing mentioned before. Approximately
half of all inherited thrombocytopenias, as well as some PFDs, are part of more complex
syndromes, in which the platelet defect is accompanied by a high probability of clinically
relevant alterations in other cell types, tissues, or organs (e.g., Wiskott–Aldrich
syndrome, Hermansky–Pudlak syndrome [HPS]), or with a predisposition for developing
a neoplastic disease (e.g., RUNX1- or ANKRD26-associated thrombocytopenia).[26]
[27]
[28] In patients in whom platelet function abnormalities are not isolated but syndromic,
an accurate and complete diagnosis that reflects the underlying molecular pathology
is important for future management.
From Candidate Screening to Whole Genome Sequencing
From Candidate Screening to Whole Genome Sequencing
When identified variants are considered as relevant and disease-causing, it is important
to perform segregation studies whenever possible. This strategy is especially useful
to confirm new variants and to improve the genotype–phenotype associations. It is,
however, not applicable to cases with a nonspecific phenotype, where there is no obvious
candidate gene, or when multiple genes are known to be causative (e.g., in cases of
GT, BSS, HPS).[13]
[29] Although GT or BSS does not essentially require genetic testing for diagnosis, it
will support to stratify genotype–phenotype associations and is indispensable for
a (co-)segregation study in parents and siblings for potential genetic counseling
to assist family planning. HPS is often diagnosed by the combination of laboratory
findings (reduced platelet count, impaired platelet function, absence of platelet
dense granules) and clinical features (oculocutaneous albinism, nystagmus).[30] Various genes have been identified to cause HPS, but dependent on the affected gene
complex symptoms can be mild, whereas some mutations harbor additional risks like
lung fibrosis.[30] The underlying gene as well as the gene variant may therefore be instrumental to
determine when and how often future clinical screenings are recommended.
With the advent of HTS at the beginning of this millennium, a more accurate diagnosis
of IPDs has become available.[31]
[32]
[33] Technical progress has made it possible to sequence preselected regions on the entire
genome (targeted sequencing, panel sequencing), coding regions (whole exome sequencing)
or the whole genome at rather low cost.[21] All sequencing approaches have several advantages and disadvantages, which have
to be considered carefully before starting genetic testing ([Table 1]). The HTS approach allows to receive information on large genes or groups of genes
with a suspected disorder like GT, BSS, or HPS, without performing numerous Sanger
sequencing reactions. There are typically groups of relevant genes analyzed depending
on the underlying functional diagnostics that can readily identify variants in genes
that are rarely analyzed. However, there are also some risks in HPS analysis: When
large gene sets are analyzed at a time, the bioinformatics might result in many variants
of unknown significance and associations could be made to quickly. The filter settings
during the analysis will provide few or many putative genes that would all require
a follow-up. For some genes, the coverage is still low; this needs to be specified
and documented on the final report and, ideally, Sanger sequencing of these gene regions
has to be performed to exclude that disease-causing variants remain undetected. Finally,
it has to be specified which genes are analyzed: (1) unrelated oncogenic or preleukemic
genes are not to be analyzed; (2) certain disease-relevant genes like RUNX1,[26]
ETV6,[34] or ANKRD26
[28] harbor an increased risk to develop leukemia. Patients (or parents) need to actively
specify whether they want these genes to be looked at and whether they want to have
this information revealed; (3) identified variants might not explain the underlying
disease of the patient, but could have information on a carrier status that could
become clinically relevant: heterozygous mutations in F8 might not explain a platelet-based bleeding phenotype, but could become relevant
when a (future) son inherited the affected gene and developed hemophilia. These options
should be discussed before the analysis is performed to avoid confusion how to report
these variants. When patients agree to perform in research-based analyses, researchers,
clinicians, and patients/parents should actively decide which information should be
shared and which should not.
Table 1
Advantages and disadvantages of different sequencing methods for genetic testing of
IPDs
|
Sanger sequencing
|
High-throughput sequencing
|
|
Targeted gene panel sequencing
|
WES
|
WGS
|
|
Advantages
|
|
Fast and cost-effective for low target number
|
Rapid and cost-effective for high target number (> 20 targets)
|
All genes analyzed in parallel
|
All genes analyzed in parallel
|
|
Established workflow
|
Analysis of disease-associated genes
|
Identification of new genes
|
Identification of new genes
|
|
Simple data analysis
|
Limited datasets
|
|
Identification of noncoding variants
|
|
Longer reads (500–700 bps)
|
Relatively inexpensive
|
|
Reliable detection of CNV
|
|
|
|
Detection of structural variants
|
|
|
|
Most uniform depth of sequencing
|
|
Disadvantages
|
|
Low number of targets
|
New genes cannot be identified
|
Large datasets
|
Very large datasets
|
|
Not as cost-effective for high number of targets (>20 targets)
|
Noncoding variants are detected only if targeted
|
Limited detection of noncoding variants
|
Relatively expensive
|
|
Insensitive to CNV
|
Difficult detection of CNV
|
Difficult detection of CNV
|
Short reads (150–300 bps)
|
|
Low discovery power
|
No detection of structural variants
|
No detection of structural variants
|
Complicated analysis
|
|
Requires updates as new disease-associated genes are discovered
|
Short reads (150–300 bps)
|
Risk of incidental findings
|
|
Short reads (150–300 bps)
|
Relatively complicated analysis
|
|
|
Risk of incidental findings
|
Risk of incidental findings
|
|
Abbreviations: bps, base pairs; CNV, copy number variation; WES, whole exome sequencing;
WGS, whole genome sequencing.
One obvious advantage of HTS is the small amount of blood needed and little risks
of preanalytical artifacts. Buccal swabs are also possible for DNA isolation. Many
research groups and consortia have already chosen HTS to accelerate the time to diagnosis
or to identify novel genes and variants involved in the pathogenesis of IPDs.[13]
[32]
[35]
[36]
[37]
[38] Importantly, HTS has to meet specific quality criteria for diagnostic means (e.g.,
a minimum coverage per base; often 20–100 reads per base).[13]
[20] Inadequate coverage can be caused by GC-rich domains (often present in the 5′ region
including the first exon), highly homologous regions, homopolymers, and repeats of
any size.[39] Low-coverage areas should thus be filled in by orthogonal technologies, such as
Sanger sequencing, to ensure adequate clinical sensitivity. Analytic achievements
could be obtained by grouping so far known and functionally similar genes for suspected
pathophysiology in one HTS gene panel. The THROMKIDplus Study Group, for example,
developed, verified, and evaluated a targeted, panel-based next-generation sequencing
approach comprising 59 genes associated with IPDs.[40]
Despite the progress in targeted HTS, a significant proportion of patients with IPDs
still fail to receive a molecular diagnosis (i.e., detected pathogenic variant), despite
being genotyped. In 2016, the THROMKIDplus panel-based sequencing approach provided
a molecular diagnosis for only 26% of patients.[40] In the recently performed “Children with Inherited Platelet disorder Surveillance”
study (CHIPS), in which 139 children with inherited thrombocytopenias were enrolled,
the defective genetic locus could be identified in only 73 children (53%).[41] These low detection rates may reflect the fact that there is currently a lack of
knowledge about all genes involved in the regulation of platelet production and function,
and that other causative defects in noncoding genomic regions or in acquired gene
defects for IPDs are overlooked by panel sequencing. One successful strategy for identifying
new IPD disease genes, as well as for molecular diagnosis in established IPD genes,
involves whole exome or whole genome sequencing, especially when combined with other
selective approaches (e.g., with combined segregation analysis by Sanger sequencing)
or complementary functional studies to ensure clinical relevance.[42]
[43]
[44]
[45] However, these analyses are still labor-intensive and require specialists with experience
in bioinformatics analysis as well as suitable functional tests to interpret the relevance
of novel variants.
Legal Basis and Ethical Aspects
Legal Basis and Ethical Aspects
The Swiss and German public health systems permit genetic analyses for diagnostic
reasons since 2015 and 2016, respectively.[40] At that time, the German health care system covered the cost of a genetic diagnosis
of up to 25 kilobases of coding sequence for patients with public health insurance,
if a special letter of referral (“Überweisungsschein Muster 10”) was provided, which
typically allowed the analysis of groups of 5 to 10 genes. As of 2021, this restriction
no longer applies, and analyses of larger panels or the entire exome to detect clinically
relevant constitutional genomic mutations in the postnatal setting are covered by
insurance (“EBM 11513—Postnatale Mutationssuche zum Nachweis oder Ausschluss einer
krankheitsrelevanten oder krankheitsauslösenden konstitutionellen genomischen Mutation”).
According to the Gendiagnostikgesetz (Genetic Diagnosis Act, GenDG), patients have
to be properly informed about important ethical aspects, which arise with genetic
testing (i.e., informed consent required). The ISTH has provided suggestions that
can be modified to fulfill national guidelines and laws.[46] Variants in IPD-related genes like RUNX1, ANKRD26, or ETV6 are associated with an increased risk for malignant disease (see the previous section).
Handling of such unsolicited findings may bear an ethical dilemma for the attending
clinician, since the patient's information self-determination is protected by the
GenDG's basic principle: the right not to know. Although good genetic testing practice
is ensured in this way, it bears the risk that pathogenic variants in IPD genes are
not allowed to be analyzed and, thus, remain unidentified and finally not reported.
These issues need to be carefully discussed with patients and parents, or with assistance
of social (pediatric) specialists, often present in the respective clinical centers.
Genetic testing may also reveal unexpected relationships between family members. Therefore,
prior to a complex genetic analysis, patients have to be asked whether they want to
receive unsolicited findings and they should be able to actively opt-in or opt-out.[46] Adequate patient education and documentation of informed consent according to the
GenDG is thus fundamental and a legal requirement. In this context, children represent
a particularly vulnerable group of patients, since they might be too young to participate
in decision making.[47]
[48] In Germany, there are recommendations regarding predictive genetic testing in minors,
that is, excluding testing of late-onset genetic disorders or carrier status (guideline
by the German Commission on Genetic Testing [GEKO] section IV.3 [Bundesgesundheitsbl 2011, 54:1257–1261, DOI: 10.1007/s00103-011-1354-6]
and guideline on genetics in minors by German Society of Human Genetics [https://doi.org/10.1007/s11825-007-0059-6]). According to the GenDG, genetic testing for differential diagnosis, in contrast,
does not have any age restriction. Of note, there is often a misunderstanding regarding
the clinical implications (where the GenDG is applicable) and (additional) biomedical
research, which is explicitly exempted from the GenDG. Here, informed consent is also
required, and studies are evaluated by a local or institutional ethical committee.
Informed consent is typically age-grouped into minors of 8 to 12 years and 13 to 17
years, where the child's concerns and wishes are integrated. In long-lasting cohorts,
a re-consent might be required once the child becomes 18 years old, especially when
biomaterial (DNA) has been preserved in laboratories or biobanks.
Analysis
All identified variants need to be further evaluated and consolidated. It is pivotal
to have “in silico filter settings” to a level that not too many variants are reported.
All variants are stratified according to the recommendation provided by the American College of Medical Genetics and Genomics (ACMG).[49] In case an already reported variant associated with the same phenotype has been
detected, the variant can readily be classified as “pathogenic” (class 5). Similar
variants (same position, another amino acid change, or a neighboring position) might
be considered as “likely pathogenic” (class 4). The biggest concern arises (1) when
a variant (or a gene) has not yet been described, (2) where a gene has been rarely
or not at all reported, (3) all evidence comes from animal studies, or (4) the overall
evidence is limited. These variants of unknown clinical significance (VUS) (class
3) require further clarification, including a co-segregation analysis with parents,
siblings, or other associated family members from the same pedigree or appropriate
functional tests. Typically, class 3 variants should not be reported or only mentioned
cautiously, to avoid that a genetic variant becomes associated with a person without
being disease causative for the patient's phenotype. These variants usually remain
in the patient file and do not only associate the patient with a nonpathogenic variant
but also curb the interest to identify the correct genetic cause. Since genetic knowledge
is expanding, a reanalysis or reinterpretation of variants is recommended after a
few years.
Variants that are known to not cause the disease are grouped as “likely benign” (class
2) or “benign” (class 1). These variants are typically under-reported, as the effort
to submit a variant to databases is less exciting when the result is negative (i.e.,
not explaining the phenotype). We highly encourage all clinicians and scientists to
make the effort to report all new variants to databases.
The NCBI-hosted ClinVar public archive has installed expert panels to collect and
continuously evaluate the evidence for genes and variants in multiple disease settings.
The Hemostasis/Thrombosis Gene Curation Panel has been fully established in 2019 and since (March 2023) has curated 103 genes.
Of note, “lumping and splitting considerations” are made to certain genes to account
for the problem that for some genes like MPL a certain set of variants will cause one disease (CAMT; OMIM #604498), while other
(mostly acquired) variants cause myeloproliferative disorders.[6]
GP1BA, for example, is causative for two distinct disorders: BSS[9] and platelet-type von Willebrand disease.[50]
[51] In ITGA2B, the “classical” autosomal recessive variants lead to GT, while other variants act
in an autosomal dominant way and cause the newly defined “platelet-type bleeding disorder
16” (OMIM #187800).[8] Genetic variants known to cause syndromic disorders might not be fully covered.
The expert panel will consider both genetic and experimental (and animal-based) evidence
according to a score sheet.[52] Finally, different classification is indicating whether the genetic evidence for
the bleeding disorder is definite, moderate, or limited. The latter might occur when there are only few case reports in the literature and
additional evidence from biological assays or mouse models is missing. Genes are continuously
reevaluated, and the class of evidence might increase, but also decrease. For some
genes like P2RY12, encoding for the platelet ADP receptor P2Y12, the entire evidence was surprisingly
only moderate. The full list can be found at https://clinicalgenome.org/affiliation/40028/.
The most comprehensive list of genes is also compiled by the ISTH in the SSC Subcommittee on Genomics in Thrombosis and Hemostasis. Here, genes are classified according to gene curation properties and stratified
into Tier 1, Tier 2, and Tier 3. Genes might move up or down between the tiers depending
on the underlying evidence. The genes with the highest importance to study are the
“GoldVariants.”[52] This data can be found at https://www.isth.org/page/GinTh_GeneLists. New variants, especially from novel genes out of whole exome studies should be reported
to this platform, so that evidence can accumulate, especially when results are not
yet published.
Final Report
When identified variants are considered to be relevant and disease-causing, it is
essential to perform segregation studies if possible (indispensable in case of class
3 [VUS] that co-segregates with other affected family members in the same pedigree,
while unaffected members do not have this variant). The identified variants might
also prompt additional functional tests that are typically performed in specialized
laboratories. It is recommended to discuss complex patients in interdisciplinary case
conferences. The final report should comprise information on the used technology and
platforms. All tested genes need to be depicted with their RefSeq (NM) numbers of
the analyzed transcripts. This allows to refer back to the exact sequence, in case
new variant updates have been published in the meantime. Some exons still have a poor
coverage in whole exome sequencing approaches, which needs to be declared in the final
report. If possible (and available), an interpretation of the identified variant in
respect to the anamnesis and functional diagnostics should be provided, otherwise
the referral to a clinician or other experts is recommended. The patient might be
informed in the setting of a human counseling (“humangenetische Beratung”) according
to §10 GenDG, but this is not mandatory.
Conclusion
IPDs, especially PFDs, are often complex disorders. Their diagnostics are overall
poorly standardized and time consuming until a final diagnosis can be made and a genetic
cause be attributed. Considering the high percentage of patients who are still left
undiagnosed or poorly classified, there is an urgent need to significantly improve
the rational approach for diagnosis of patients with suspected IPD. HTS allows to
elucidate the cause of IPDs on the molecular level. Based on this, databases are continuously
expanding and assisting to identify (non)-pathogenic variants. Future research on
IPD genes will improve the diagnosis and treatment from which finally the IPD patients'
quality of life will benefit.
