The overarching achievement is the creation of an imperfect but agile and iteratively improving, diagnostic platform within the public health service, that is aligned to unmet need, and that supports equitable state-wide diagnostic health care provision through the integration of genomic diagnostics. Notably, this improved in-house causative monogenic detection rate from a historical service baseline of 10 – 30 % for this heterogeneous and diagnostically challenging clinical cohort. The RUDDS is consistent with fundamental tenants of clinical genetic service delivery, including excellent phenotyping, partnership with genetic pathologists and laboratory staff and alignment to the patient journey. Primary outcomes of the RUDDS included:
Firstly, that it was delivered in a patient-centric manner resonant with the patient journey, specifically that as it is not limited to the duration of a research project, nor to a static consent process, that patients can engage with or leave the platform at various stages and at times relevant to them, providing news they can use and when they are ready to receive it. Additionally, through provision in a state-wide service, including outreach clinics, it could be provided with geographic equity. Secondly, by integration in clinical service , multiple options including monogenic and genomic (targeted and whole exome) analysis, non-gene based testing, and matchmaking could be offered. Thirdly, it was synchronous with precision phenotyping methods, including that 3-D facial analysis could be provided. Even if it was “just” the provision of a 3-D image that could be interrogated at clinical case review to provide an increased depth of information compared to 2D images. Fourthly, it captured new knowledge, including multiple expert reviews, specifically the documentation and process of team discussions around each individual either before and/ or after testing provided opportunities for collective up-skilling with likely future benefit to future families. For instance, multiple cases with clinical indicators of possible mosaicism and normal testing on peripheral blood, altered clinical practice for tissue selection in subsequent cases, and lead to new processes for acquiring the relevant samples. Additionally, the new knowledge of the mTOR mutation offered a new alternative for seizure treatment. Fifthly, integration with current translational genomic research activities, specifically, but not limited to the the SeqNextGen project described above. In some instances, patients that could not be resolved with the SeqNextGen project were recruited into research projects with laboratories with specific phenotype domain expertise. Also, as a by-product of the relationships established through this collaboration, and following the case of a familial mTOR described above, the first genomic reference range data for Aboriginal Australians [12] was translated from research to implementation as de-identified frequency data in our clinical laboratory; Sixthly) flexibility for integration with further clinical research including additional diagnostic approaches, knowledge management platforms, community engagement, policy and models of care. This includes: i) whole genome sequencing; ii) the planned implementation of a knowledge management platforms for mining of free text, case sharing, discussion and an application programming interface to push data to Matchmaker Exchange, namely Patient Archive; iii) community engagement, such as that enabled through the delineation of a critical mass of diagnosed and undiagnosed patients and relatedly by directly partnering with community peak bodies to develop policy and models of care. Finally, less tangible outcomes included: increased workplace satisfaction for staff though improved access to molecular confirmation, as indicated by repetitive informal feedback, and noting that this was not assessed through formal surveys during the period described; and the potential for professional development through the shared experiences and quality improvements resulting from definitive diagnoses.
The overall mutation detection rate was in accordance with expectation at 30 %. Similar to other reports [8–10], the diagnostic yield was proportional to the specificity of the described phenotype, including affectation of multiple organ systems, rather than a single organ system. This was reflected in the following mutation detection rates hereditary haemorrhagic telangiectasiae (5/5, 100 %), RASopathies and overgrowth disorders (7/16, 44 %) and presumed X-linked mental retardation (0/7, 0 %).
Two illustrative cases are described. Firstly, over a 10 year period an Aboriginal family from remote Australia had been seen in an outreach clinic. Multiple siblings had the same unique phenotype characterised by intellectual disability, macrocephaly, small thoraces, asymmetric visceral overgrowth, connective tissue dysplasia, predisposition to infection, and variably one to multiple cafe-au-lait lesions, hemi-megalencephaly with perisylvian polymicrogyria. Over the decade, multiple diagnoses had been considered and many monogenic tests had been performed with normal results, largely to investigate RASopathies and overgrowth disorders. Guided by the phenotype, WES was targeted to genes known to be associated with the two presumptive diagnostic groups and overlapping genes, including mTOR. A likely pathogenic variant in mTOR was identified. The mutation segregated with disease in the family. Following a protracted search for a suitable laboratory, functional studies were performed that supported a gain of function mechanism for this gene [13]. This finding also supported the possibility of mTOR inhibitor drug repurposing. This case highlights the importance of phenotyping; the need for coordinated access to functional studies, such as in the first instance might be enabled by a virtual network of functional analyses; and management implications of a molecularly confirmed diagnosis.
In the second example, a child presented with features consistent with megalencephaly-capillary malformation syndrome; specifically, prenatal onset impressive macrocephaly, infantile onset hydrocephalus, capillary malformations and connective tissue dysplasia. Targeted exome sequencing of PIK3CA and related genes was performed on DNA extracted from peripheral blood and skin fibroblasts; a mosaic mutation was detected in the latter but not the former and was subsequently confirmed by Sanger sequencing of DNA from skin fibroblasts and cerebrospinal fluid. This case again highlights the importance of phenotyping to guide genomic analysis and highlights the selection of an appropriate tissue for testing.
Provision of this service within a public health setting and with multi-expert review was an approach that was tailored to local circumstances, including optimal use of limited health resources targeted to the unmet need of a population where it is was most likely to have immediate clinical utility and deliver patient benefits. Key to the patient-centric delivery of this service was the pivotal involvement of genetic counsellors at multiple points in the diagnostic pipeline and for ongoing support as required. Other critical success factors included partnerships with translational research to augment future diagnostic analytic capacity, promotion of access to functional analysis, and the initiation of economic analyses and studies of patient and family experience, as well as the development of improved models of care. Similarly, active community engagement was informative and supportive at all stages. Finally, the involvement of a public health genomic policy unit supported implementation and sustainability.
By iterating within the clinical service, known or unanticipated real-world bottlenecks can be identified and pragmatically addressed. Also, real-time in-service studies of health care provision can be performed and are in progress. Process improvements to address identified bottlenecks will need to include more timely ascertainment of human phenotype ontology terms in a manner aligned to clinical flow [14], as this will streamline communications with the clinical laboratory and facilitate the better use of phenotypic information in the analytic workflow. Also, the need for capacity building including promoting the development of a workforce that is knowledgeable of the evolving relative place and limitations of genomic and non-genomic investigations. Additionally, given the significant proportion of cases that remained undiagnosed, the need for complementary approaches for diagnostically intractable cases, including local and international undiagnosed diseases programs [15], was reinforced.
A limitation of this work which might also be viewed as a critical success factor, was the lack of internal consistency of some technical instrumentation and analytic tools used during the evaluation of this cohort. This was reflective of and responsive to changes in technology and was a testament to the endeavour and flexibility of the laboratory team that delivered within the heterogeneity and imperfection of clinical process. What was also demonstrated very clearly was that it was the phenotype-informed and patient-centred processes that were instrumental to applying technology to enable improved diagnostic yield.