We have performed a nation-wide audit of VUS recorded in the clinical databases of Australian familial cancer services to determine the scale of unrecognised actionability, and to map the current and required processes to achieve effective variant review. Our ultimate aim is to inform future improvements in Australian practices concerning variant classification, reducing variant uncertainty and improving patient clinical management.
With a straightforward crosscheck of the 944 clinically detected VUS retrieved from FCCs against ClinVar, we found that a notable number of the historical reports recorded in FCC internal clinical databases had not undergone updates and/or reassessment with existing new evidence. Approximately 20% of the so-called VUS had an Expert Panel submission as P/LP or B/LB, with an additional 5% exhibiting multiple non-conflicting submissions in the same categories, information that was returned to the FCCs. Most of these ClinVar submissions were in the benign direction, however there are many reasons to consider variant downgrades as clinically important, including the opportunity to avoid unnecessary clinical risk management interventions or relief from any continuing anxiety, although it can also be an indication to continue searching for alternative genetic explanations. According to the results from the survey of FCCs, the majority of reclassifications in either direction, specifically 100% of VUS upgrades and 67% of downgrades, would be communicated to patients. This attitude aligns with findings from a US-based survey study where both oncologists and genetic counsellors concurred that all reclassified variants, even when they do not represent a change in clinical management, should be disclosed to patients (22). Of the greatest concern though were the 11 variants with Expert Panel submissions as P/LP, alongside nine variants with multiple non-conflicting P/LP submissions (collectively corresponding to 26 patients), which have a more clear-cut implication for the management of the patients and their family members.
For the large group of remaining variants that are VUS, conflicting in ClinVar or absent from the database altogether, we found that application of gene-specific guidelines developed by ClinGen Expert Panels resulted in a 79% reduction in the VUS rate, even when relying solely on publicly available information. Of the 698 variants in the “Other” ClinVar group (as at 1 September 2022), 3% (19 variants) could be reclassified as P/LP and 77% (534) as B/LB by the application of gene-specific guidelines. When added to the variants with Expert Panel or multiple non-conflicting submissions, our analysis suggests that a total of 4% of the initially provided 944 unique VUS (39 variants), could potentially be reclassified as P/LP, while more than three quarters of the VUS (745 out of 944, 79%) could be reclassified as B/LB utilising this approach.
To better understand the factors contributing to these findings, and to describe the barriers to FCCs and laboratory identifying and implementing reclassifications, we surveyed both groups. Representatives from a large majority of FCCs (89%) indicated that they do not routinely update their internal variant database as they are generally not aware of any potential reclassifications. On the other hand, in the laboratory survey results, laboratories indicated that they notify the FCCs of variant reclassifications in nearly all instances, with the primary reason prompting VUS review being an FCC request to the laboratory for additional or updated information. This highlights a major flaw in the current VUS review process: FCCs heavily rely on laboratories to keep them updated of variant reclassifications, as they do not have the clinical capacity and/or expertise to do their own reviews, but laboratories rarely initiate VUS review without a clinical prompt. The results suggest that only a relatively low proportion of VUS found in patients will be put forward for further review due to a request from the FCC or as a result of a relative undergoing testing, and that in most instances where a potential reclassification could occur based on updated information and/or classification methods, this opportunity is missed. Only two of the six surveyed laboratories indicated that they initiate their own variant re-review on a regular basis. Resources was the unanimous reason for no more regular or initiated variant re-review, and all but one of the laboratories noted that they would prioritise VUS review if they had more funding to do so. One laboratory representative elaborated that regular VUS review, regardless of the funding, requires significant consideration of the downstream clinical interactions and interventions. There are possible legal implications if there have been active clinical interventions due to classification as P/LP, such as prophylactic mastectomy and oophorectomy, with subsequent downgrades, or clinical inaction contributing even to deaths due to classifications as LB or VUS with subsequent upgrades, even within the same families.
These results are a prompt to consider how a more effective and reliable process could be implemented for identifying reclassifiable VUS in clinical datasets and communicating updated curations to facilitate optimal care for patients. A simple and affordable recommendation to streamline the VUS review process between clinical services and laboratories is to promote automated notifications to the laboratories of Expert Panel reclassifications for variants within their curation system. In Australia, there is potential to facilitate this by leveraging the properties of Shariant (23), a system for real-time sharing of variant data between Australian FCCs and laboratories, which has recently introduced this capability. This could be supplemented by general look-ups to ClinVar, preferably in an automated manner as recently proposed (24). Ideally, this approach would be coupled with encouraging proactive notification of reclassifications by laboratories to FCCs. However, Expert Panels do not typically perform large-scale reviews of variants, as this typically exceeds the capacity of the VCEPs and is not prioritised by ClinGen, which usually focuses on variants with multiple uncertain or conflicting submissions. Therefore, while this "ClinVar notification" approach can be beneficial, it will not address the larger number of variants that are not reviewed by Expert Panels but would still benefit from more frequent laboratory-based reviews following gene-specific ACMG/AMP guidelines.
There was no clear consensus in the survey responses around which group bears the final responsibility for the task of continuing review of VUS to identify potential reclassifications. A more comprehensive solution is required for promoting periodic laboratory review of VUS for potential increased actionability. Overall, our VUS review work revealed a potential pathogenic reclassification for over 4% of unique variants in selected hereditary cancer genes currently recorded in clinical databases as VUS. We expect that this reclassification rate would be greater with inclusion of evidence held privately within laboratories. The reclassification rate is also likely to increase with evolution of ClinGen gene-specific classification guidelines, which has been demonstrated in several VCEP-related studies to reduce the VUS rate (25–27).
Regular VUS review and variant reclassification will require resources, adding to the routine work of trained staff. Numerous studies have demonstrated the cost-effectiveness of genetic testing in women with breast cancer (28–30). In response, countries such as Australia and the UK have integrated genetic testing for hereditary breast cancer risk into their respective government funded health services (Medicare and the NHS), targeting patients meeting specific eligibility criteria, including those with a 10% risk of carrying a germline pathogenic variant based on clinical presentation. It should be noted that VUS review and variant reclassification activities to identify a P/LP variant should be less costly than the initial genetic testing that incurred the costs of the initial consultation, sample collection, laboratory analysis, sequencing bioinformatics. Costs associated with VUS review would be limited to resources for re-curation costs, and re-issue of reports when necessary. Additionally, reclassifying VUS as B/LB (79% reclassification rate as per our work) has the potential to lead to savings by avoiding the cost of unnecessary clinical interventions, recognising that not all VUS are considered clinically relevant and management might be dependent on other factors, and by removing variants from the list of VUS requiring ongoing review by the laboratory. Further, for VUS previously returned to patients, there is potential to provide reassurance and relieve anxiety following a B/LB reclassification. Therefore, publicly funding the VUS review process in countries where hereditary cancer testing is already publicly funded appears likely to have economic benefits. In terms of VUS review frequency, two of the laboratories reported conducting this activity every 1–2 years, aligning with reported most effective frequencies (31, 32). The inconsistency in approaches between diagnostic laboratories reveals a need to develop a national approach to standardise variant reclassification protocols. The consensus framework on variant reclassification developed by CanVIG-UK (33) might provide a useful starting point in this regard.
Future directions of this work include assessing the clinical impact of our findings, and the cost-effectiveness of implementing routine variant review and reclassification practices. In parallel, national efforts should be directed towards promoting collaboration between FCCs and laboratories to ensure timely notification processes, potentially through standardised protocols and increased communication channels.