Genomics Vs. Haematology and Transfusion Laboratory

Genomics, the study of the complete genetic sequence of an organism, is playing an increasing role in clinical medicine. This change has been enabled by the rapid development of new technologies, which can sequence a human genome within a day at a cost of less than £700. Initially the integration of genomics into healthcare has focused on the diagnosis of rare inherited diseases and the management of cancer1. However, it is hoped that genomic medicine will expand into other fields, meaning all specialties will need to adapt. It is important to consider what changes will be necessary, so we can begin developing new services and training programmes. For example, will the next generation of haematologists need to learn how to analyse blood films, or would the time be better spent learning the coding languages necessary for genomic analysis? Indeed, will any of the current techniques utilised in the haematology and transfusion laboratory survive the genomic revolution? Or will we come to rely entirely on genomics for all diagnostic, prognostic and therapeutic queries?

For genomics to supersede the traditional haematological techniques the following conditions must be met. Firstly, genomics must be superior, or at the very least non-inferior, to currently used tests. This discussion will comprise the bulk of the essay, and include an outline of current haematological tests, the genomic techniques which may replace them and the advantages and disadvantages of the latter. Secondly, the transition to genomic medicine would have to be cost effective, environmentally sustainable and allow equitable healthcare.

The Current State of Affairs

Before discussing genomics, it is important to establish the scope of the tests that would be replaced. Consider a patient with chronic myeloid leukaemia (CML). The first clue to the diagnosis may be leucocytosis on the full blood count, which microscopic analysis of the blood film would reveal to be mostly mature granulocytes. From here, the key test would be RT-PCR or fluorescent in-situ hybridisation to establish the presence of the BCR-ABL oncogene which is pathognomonic for CML. Further genetic testing aids in determining prognosis, namely obtaining a bone marrow aspirate for karyotyping to assess for the presence of cytogenetic changes associated with progression to the blast phase. If this patient required a blood transfusion, serological tests would be used to determine their ABO and RhD groups. The above journey is reflective of other haematological malignancies, many of whom have characteristic mutations with diagnostic and prognostic value. Similarly, genetic testing is also utilised in the diagnosis of inherited haematological conditions, such as sickle cell anaemia8.

Genomics: An Overview

For genomics the patient journey would begin similarly, with the collection of blood or other tissue sample. These samples need to be fresh frozen to allow extraction of high quality DNA1, which can then be amplified and sequenced using next generation sequencing (NGS) technologies. NGS involves sequencing multiple fragments of DNA simultaneously, the overlap between these fragments is then used to generate a complete genome sequence. Once a sequence is obtained, the extensive process of analysis can begin. For patients with cancer, this would involve comparing the genomes of cancerous and healthy cells to identify mutations known or predicted to have an effect on prognosis and treatment response.

It is worth acknowledging that genomics is only one branch of functional genomics. The latter field is devoted to the study of how genes and their products function to produce a phenotype in a cell or organism. It comprises many different "-omics", including transcriptomics, metabolomics and proteomics. Given these other "-omics" require additional techniques and are less well established clinically, for the purpose of this essay the capacity of genomics alone to replace the haematology and transfusion laboratory shall be considered.

Advantages of Genomics

The principal advantage of genomics is the sheer amount of information that can be obtained. For example, malignant cells typically have a large number of mutations, some of which determine diagnosis, prognosis and response to treatment. This is exemplified by Ph-like acute lymphoblastic leukaemia (ALL). Unlike many other subgroups of ALL, Ph-like ALL is not defined by a founding chromosomal rearrangement. Instead, the genome is peppered with genetic alterations in cytokine receptors and tyrosine kinases. Whilst this subgroup currently has an overall poor prognosis, these multiple mutations allow for the possibility of targeted therapy. Thus, genomic assessment of patients with possible ALL could give a rapid and accurate diagnosis of Ph-like ALL and other subgroups, personalised prognostication and targeted therapy in a single test. As a result, genetic tests currently used in the haematology laboratory, such as PCR and karyotyping, may become obsolete.

Another advantage of genomics is that it may be used to assess molecular heterogeneity. It is well-established that an individual's cancer will consist of multiple subclones, derived from a founding clone but with divergent mutations. These clonal populations can be precisely defined using genomics coupled with deep sequencing or single-cell genomics. Utilising these techniques early in treatment may identify clones which are naturally more resistant to therapy and are thus likely to become the predominant clone at relapse. As cancer is a dynamic disease, serial genomic sampling may also demonstrate shifts in clonal populations and the development of new mutations, thereby enabling changes in treatment regimens before a new clone becomes established.

The advent of the genomic era is also likely to accelerate research. It is envisioned that the data generated from genomic sequencing, along with detailed clinical information, will be available to researchers. Analysis of such a database may reveal new patterns of mutations associated with disease outcome or pathogenesis. For example, the genetic basis of ALL remains undetermined in a significant minority of patients. Defining the characteristic mutations in these cases could lead to the identification of new treatment targets and subsequently improved survival rates. Another exciting area of research is using genomic sequencing of cell-free DNA in the bloodstream to identify circulating tumour DNA (ctDNA) as a method to monitor response to treatment or screen for malignancy. For example, in a 2012 study whole genome sequencing (WGS) of cell-free DNA of 1002 individuals identified four cases of lymphoma and one of myelodysplastic syndrome with excess blasts1. In the future, cancers may be diagnosed, monitored and treatment decisions made using genomic analysis of ctDNA alone11, reducing the need for more invasive testing.

Another area in which genomics is already proving to be useful is in the identification of disease-causing mutations in rare diseases. This application would be valuable in the management of haematological diseases thought to be the result of as yet unidentified mutations. For example, 2% of patients with a bleeding disorder do not show abnormalities on laboratory testing. It is thought that these patients are likely to have defects in their platelets or vessel walls12, which may be identified with genomic testing. Thus, genomics may supersede current clotting and platelet function tests by identifying both established and novel disease-causing mutations in patients with bleeding disorders.

There are additional advantages to using genomics over specific genetic tests in the diagnosis of inherited haematological conditions. For example, WGS in a patient with possible sickle cell anaemia could not only identify the disease-causing mutation, but may also be used to screen for genetic variations which render an individual more susceptible to specific complications such as osteomyelitis. Similarly, genomes can be screened for markers which indicate the effectiveness of medications and the susceptibility to side effects. This includes genetic variants associated with opioid sensitivity, potentially allowing more effective analgesia in sickle cell crises or haematological malignancies. The hope would be to produce more tailored care, resulting in more efficient allocation of resources and reducing unnecessary suffering.

Regarding transfusions, genetics is already an integral part of blood grouping and is replacing traditional serological tests in some institutions. Indeed, genotyping may be superior as it can identify antigens expressed at a level too low to be detected by a serological test, or for which such an assay is not available. Therefore, genetic testing enables extended grouping, allowing for more accurate donor-recipient matching to prevent alloimmunisation. In turn, these genetic tests may be superseded by WGS, which could be used to group all red blood cell (RBC) antigens.

Disadvantages of Genomics

The most significant limitation of genomic testing is that not all diseases are genetically determined. Whilst some, such as sickle cell anaemia, are determined by a single inherited gene mutation8, others are purely environmental, for example anaemia resulting from dietary B12 deficiency. In reality most diseases fall somewhere between these extremes, with genetic and environmental factors influencing development and severity of disease. Thus, whilst genomics may be used to identify those at increased risk of a specific condition, proving that an individual is currently affected will likely require additional testing. For example, to diagnose an individual with autoimmune haemolytic anaemia (AIHA), a test is needed to demonstrate the presence of antibodies directed against a patient's own RBCs. NGS techniques can be used to examine the antibody repertoire, however this is an incredibly complicated process. Antibody production requires B cells to rearrange and self-mutate their own genomes. As a result, each group of B cell clones has a unique genome producing a unique antibody. Assessing such incredible diversity typically requires a mix of both bulk and single cell sequencing. Even then, genomic sequences of B cells in peripheral blood are not a perfect representation of antibodies present in sera. Thus the current test of AIHA, the direct antiglobulin test, is both simpler and more sensitive than genomic sequencing.

The limited capacity for genomics to replace antibody-based tests also has significant implications for transfusions. As discussed above, genomics could allow for more accurate matching for blood transfusions. However, given the numerous RBC antigens and limited blood supply, it will be necessary to transfuse blood where there is donor-recipient mismatch for more minor RBC antigens, including those with alloimmunisation potential. Whilst only a minority of patients will develop detectable alloantibodies following transfusion, these can cause delayed, or more rarely acute, haemolytic reactions. Therefore, serological testing to identify the development of alloantibodies following transfusion will continue to be required.

Another significant disadvantage of genomic testing is speed. WGS is significantly faster than two decades ago1, however it still takes approximately a day to obtain a sequence and this does not include time for sample collection, DNA extraction and analysis. As a result, WGS may not be appropriate for emergency situations. For example, blood is often required urgently and whilst O-negative blood can be safely given, it is a valuable resource with limited supply. In such cases, it seems likely that there will be a continuing need for crossmatching, which can be in performed in under an hour.

Furthermore, whilst information gathering is one advantage of genomic testing, it is not without its drawbacks. Haematologists risk being overwhelmed by data of unclear clinical utility, a limitation demonstrated most clearly by variants of unknown significance (VUS). When genetic variations are identified in the genome they are classified as benign or pathogenic based on evidence such as the predicted effect of the mutation or whether it has been seen previously in individuals with the disease. However, this classification is often ambiguous, resulting in VUS variants which cannot be confirmed as benign or pathogenic. These are a considerable source of frustration for both patients and doctors. Patients may worry unnecessarily about variants which are in fact harmless. For doctors, ignoring a VUS may mean missing a mutation which is key to pathogenesis, alternatively further investigations may waste considerable time and resources determining that a variant is benign.

Similarly, the untargeted nature of genomics means that incidental, or secondary, findings will be commonplace. These are variants which have no impact on the primary diagnosis, but may, for example, indicate an increased risk of developing an unrelated condition. Proponents argue these findings are beneficial as they could allow for preventative therapy. However, this relies on such treatment being available and efficacious in patients identified as at-risk based on genetic profile. Moreover, some patients may not want to be informed of secondary findings, which raises serious ethical and practical questions as to how we can ensure that their data will remain secure and when a disclosure should be made.

Other Considerations: Sustainability and Equality

The switch to genomic medicine will require significant financial investment. In addition to the cost of the sequencing itself, new infrastructure will be needed to transport and process clinical samples, analyse and store data, and train healthcare professionals. Moreover, in a world increasingly affected by climate change, it is imperative that we consider the environmental impact of a switch to genomics. Genomic data is not a carbon neutral asset. Storing and processing such vast quantities of information requires a large and constant supply of energy which typically comes from non-renewable sources.

In cases where genomics can be used to save lives, such as enabling personalised cancer therapy, the environmental and financial cost seems justifiable. Indeed, it may be more cost-effective as genomics could reduce time to diagnosis and prevent unnecessary treatment by providing targeted therapy. Conversely, in other cases the benefits may be more marginal and we need to consider whether this is a worthwhile allocation of limited resources. For example, as discussed above, a pharmacogenomic profile could be used to inform pain management during a sickle cell crisis. However, genetically determined sensitivity to analgesics and their side effects is only one factor which will influence patient response, environmental and developmental factors will also play a role. As such, it may be more effective to invest in additional specialist nursing staff to allow regular monitoring of pain, timely delivery of analgesia and assessment of side effects.

Another key factor to consider is whether genomic medicine will be equitable. A major problem with the current genomic databases is that they are dominated by data from individuals of European descent. Therefore, it is more difficult to determine the significance of genetic variations and the risk of developing particular diseases in non-white individuals. Until these problems are addressed, relying on genomic medicine may result in a poorer quality of care for non-white populations and thus further entrench existing inequalities.

Conclusion

Overall, genomics is a tool to be incorporated into, not replace, the existing haematology and transfusion laboratory. There are clearly significant advantages to WGS in certain contexts, particularly for cancer where the ability to rapidly screen the complete genome in a single test will likely replace many of the current genetic tests. However, we need to be cautious not to generate excess genomic information unnecessarily. This is not only environmentally and financially costly, but risks wasting limited clinical resources on VUS or the identification of unwanted secondary findings. Moreover, we are more than the product of our genomes and most diseases arise due to complex interactions between our environments and genetics. Thus, the haematologist will continue to rely upon traditional genetic and non-genetic tests. Ultimately, WGS is like any other investigation, in that it should be used selectively in the right patients in the appropriate clinical scenario.