Abstract

Introduction

Testing of patients with cancer for high penetrance breast-ovarian cancer susceptibility gene (CSGs) BRCA1, BRCA2 and PALB2 offers three potential benefits. First, identification of a germline pathogenic variant can provide insights into the oncogenesis of their cancer, potentially informing selection of chemotherapeutic agents, including platinum and targeted therapies such as poly-ADP ribose polymerase inhibitors. Second, knowledge of elevated risk of subsequent breast, ovarian and other cancers may direct the patient towards risk-reducing surgery and/or intensive surveillance. Third, cascade testing for the pathogenic variant enables identification of family members who carry it, are at elevated risk of cancer and might benefit from medical interventions, as well as providing reassurance for those who do not carry the familial variant.¹

Historically, variant scanning along the length of a CSG was labour-intensive and thus expensive, typically taking several months in a diagnostic laboratory. Hence, germline genetic testing was largely divorced from acute diagnostic oncology, instead being initiated more typically by those who had successfully completed treatment for a prior cancer diagnosis and unaffected relatives concerned by their family history. With the advent of next-generation sequencing (NGS) technologies, sequencing of CSGs has become relatively cheap, rapid and high-throughput. Furthermore, we have good knowledge of the pathogenicity of variants in well-characterised genes such as BRCA1, BRCA2, MLH1 and MSH2, such that interpretation can be streamlined via automated bioinformatics pipelines, with a low rate of variants of uncertain significance (VUS). Thus, it is increasingly feasible from the laboratory perspective that large-scale, rapid CSG analysis could be offered routinely as part of the diagnostic workup for all patients with relevant cancers. This in principle offers opportunity for patient’s primary surgery to combine treatment of the current cancer with risk reduction for future cancers. For example, bilateral mastectomy rather than localised excision may be performed in a BRCA1-positive woman with newly diagnosed breast cancer.

However, while technological advances have driven massive improvement in laboratory capacity, substantial barriers within the upstream and downstream clinical pathways remain. The traditional model of individualised patient referral to clinical genetics for management of pretest genetic counselling, consenting, sample acquisition and return of results reflects an era in which patient volumes were low and timescales unpressured. To facilitate rapid delivery to larger populations of patients with cancer, there have been attempts to transition germline genetic testing across into mainstream oncology.^2–4 For ovarian cancer, of which there are ~7500 cases per year in the UK and a ~15% frequency of germline pathogenic variants of BRCA1/BRCA2,^{5 6} over the last 5 years there has been relatively successful implementation of universal testing using a variety of ‘mainstreaming’ models.^{2 7} For breast cancer, the incidence is much higher (~56 000/year)⁸ while the pick-up rate of pathogenic variants is more modest (~3%–5% in total for BRCA1/BRCA2/PALB2).^9–11 Delivery of germline genetic testing by mainstream breast cancer oncological clinicians has been piloted, but success has been more limited. Lack of requisite expertise, high workload, large patient numbers and perceived relevance of germline testing for immediate clinical decision-making have been cited as causes for clinician reluctance.¹²

Limitations of the precision oncology model in improving outcomes in those diagnosed with advanced cancers has led to renewed focus on improving cancer early detection and prevention. Such interventions are most impactful applied to those at very high priori risk of cancer. However; despite being the earliest, most common and arguably most widely established paradigm of high penetrance cancer susceptibility, recent analyses suggest low ascertainment, with fewer than 3% of BRCA1/BRCA2 heterozygotes across Greater London identified.¹³

Recent health economic analyses, using UK-specific and other costing parameters, demonstrate testing of BRCA1/BRCA2 genes in unselected patients with breast cancer to be cost-effective.^14–17 However, the current NHS eligibility criteria exclude >80% of patients with breast cancer from germline genetic testing, a restriction now arguably driven more by capacity and costs of clinical manpower than laboratory assays.^18–20 It is paradoxical therefore, that complex evaluation of family history to exclude ineligible patients still occupies a substantial proportion of capacity of expert genetics clinicians of which the system is so short. Furthermore, due to small family size, male transmission, fractured transmission of familial information and variation in penetrance and chance, the current family based criteria fail to catch about half of BRCA heterozygotes.^{21 22}

An additional consequence of this high threshold for NHS germline genetic testing is diversion of ‘ineligible’ patients with breast cancer to private and direct-to-consumer testing. As well as driving inequity, these laboratories function outside of the regulatory standards and informatics systems unifying the UK NHS diagnostic laboratory network, through which we ensure that variant information and classifications are consistent, shared and updated. These parallel systems have ethical implications regarding equity of access to downstream NHS-funded interventions,^{23 24} and create additional friction regarding NHS evaluation of spurious results for patients who were ineligible for NHS testing.^{24 25}

We and others have hypothesised that integration within NHS diagnostic cancer pathways of simple technology platforms could mitigate the impasse.^26–28 Requirement for detailed individualised genetic counselling to inform the decision to undertake a genetic test comes from early models for Huntington disease and prenatal scenarios.²⁹ For a cancer patient, a germline genetic test is arguably one component of a suite of tests potentially informing their cancer management, and while there are also important considerations to be made about possible implications of the result for their future and family, the information relating to the test is largely generic. Likewise, operational management of consent, sample transmission and return of results is largely formulaic. However, integration within NHS clinical and laboratory information systems of this digital pathway is critical to ensure (i) full communication of results across clinical organisations, (ii) appropriate expert management of nuanced scenarios such as uncertain variants and BRCA-negative patients with strong/unusual family histories and (iii) ongoing VUS review with patient recontact.

Over the last decade, there has been considerable focus in NHS strategy on application of digital solutions to extend and improve access to healthcare, a trend dramatically catalysed by the COVID-19 pandemic. Furthermore, there has been concurrent high priority within NHS England for expanded application of genetic testing for personalised prediction.^{30 31} We therefore sought to design a digital pathway that was rapid, patient-centred, ‘light-touch’ for clinicians, and integrated into NHS clinical, laboratory and informatics systems, by which genetic testing could be delivered to mainstream patients with cancer at potentially much greater scale. For our exemplar clinical scenario, we applied this pathway to BRCA-testing (BRCA1/BRCA2/PALB2 gene testing) in unselected mainstream patients with breast cancer.

While a digital pathway would be presumed to improve capacity and efficiency, questions remain regarding whether a digital route might adversely impact anxiety, satisfaction and/or understanding of the genetic testing process for patients with cancer. Furthermore, while we presuppose that patients with cancer would value the rapidity, convenience and flexibility of a digital pathway, these trade-offs have not been well explored, in particular for UK NHS patients.^{28 32 33} As well as evaluating patient responses to the digital pathway, we also leveraged this opportunity to compare by randomisation two approaches to delivery of pretest information. Half the patients underwent the ‘fully digital pathway’ with pretest information delivered digitally, while half the patients had a pretest telephone consultation with a genetics professional (the ‘partially digital pathway’).

We here present data on development of the BRCA-DIRECT pathway, evaluation of our internal pilot of the BRCA-DIRECT pathway in 130 unselected NHS patients with breast cancer, preliminary data from randomised comparison in this population between digital and telephone consultation delivery of pretest information and assessment of progression criterion established to support continuation of the study to the full recruitment target for well powered analyses of outcomes.

Methods

We describe materials and methods relating to three constituent activities, namely: (i) development of the BRCA-DIRECT digital pathway, (ii) evaluation of the BRCA-DIRECT digital pathway, (iii) randomised evaluation of delivery pretest information, comparing digital delivery (fully digital pathway) with telephone consultation with a genetics professional (partially digital pathway) as well as (iv) assessment against established progression criteria for continuation of the study (figure 1). For full details, see online supplemental methods.

Open in a separate window

Figure 1

The BRCA-DIRECT full study pathway, including COVID-19 adaptations to minimise aerosol generating procedures (stripped grey). Light grey: study-specific procedures, including enrolment activities and study questionnaires. Light green: BRCA-DIRECT digital pathway core activities. Dark green: fully digital pathway with delivery of pretest information digitally. Blue: partially digital comparator arm with delivery of pretest information via telephone consultation with a genetics professional. VUS, variants of uncertain significance.

Results

Study population characteristics

Recruitment to the BRCA-DIRECT internal pilot took place between 5 July 2021 and 10 October 2021 (97 days). During this time, 146 women with breast cancer expressed positive interest in participating, with 130 (89.0%) returning study consent forms and samples (figure 2; online supplemental figure 3).

Open in a separate window

Figure 2

Consolidated Standards of Reporting Trials flow chart detailing patient progression through the BRCA-DIRECT pilot study, including number of patients included in analysis at each stage, separated by pretest information randomisation allocation following enrolment. Light green: study-specific outcome measures. Dark green: digital pretest information (fully digital arm). Dark blue: telephone pretest information (partially digital arm). Light blue: standard pathway procedures. See online supplemental figure 3 for more detail on patient progression and reasons for withdrawal or exclusions at each stage.

Of the 130 women who consented to the study, 52.3% of patients were newly diagnosed and presurgical, 28.5% postsurgical under active treatment and the remainder under follow-up (6.9%) or metastatic (12.3%). Patients ranged from 33 to 87 years of age, with a mean age of 59 years old. See further demographics in table 1.

Table 1

BRCA-DIRECT pilot study patient demographics

Demographic	Groups	All	Pretest information allocation
			Digital	Telephone
		N (%)	N (%)	N (%)
Age (years) (n=129)	18–30	0 (0.0)	0 (0.0)	0 (0.0)
	31–40	4 (3.1)	2 (2.9)	2 (3.4)
	41–50	27 (20.9)	13 (18.8)	14 (23.7)
	51–60	47 (36.4)	19 (27.5)	28 (47.5)
	61–70	25 (19.4)	15 (21.7)	10 (16.9)
	71–80	21 (16.3)	15 (21.7)	5 (8.5)
	81+	5 (3.9)	5 (7.2)	0 (0.0)
Ethnicity (n=128)	Asian or Asian British	8 (6.3)	5 (7.9)	3 (4.6)
	Black African, Caribbean or black British	3 (2.3)	0 (0.0)	3 (4.6)
	Mixed or multiple ethnic groups	7 (5.5)	3 (4.8)	4 (6.2)
	Other ethnic group	2 (1.6)	1 (1.6)	1 (1.5)
	Prefer not to say	2 (1.6)	2 (3.2)	0 (0.0)
	White	106 (82.8)	52 (82.5)	54 (83.1)
Highest education (n=128)	Higher degree level	31 (24.2)	13 (20.6)	18 (27.7)
	Degree level	39 (30.5)	21 (33.3)	18 (27.7)
	NVQ or equivalent	16 (12.5)	7 (11.1)	9 (13.8)
	A-levels	11 (8.6)	6 (9.5)	5 (7.7)
	GCSE or equivalent	23 (18.0)	12 (19.0)	11 (16.9)
	No qualification	4 (3.1)	2 (3.2)	2 (3.1)
	Prefer not to say	4 (3.1)	2 (3.2)	2 (3.1)
Marriage status (n=127)	Married or partnered	82 (64.6)	41 (66.1)	41 (63.1)
	Widowed	11 (8.7)	5 (8.1)	6 (9.2)
	Single	12 (9.4)	7 (11.3)	5 (7.7)
	Divorced	19 (15.0)	7 (11.3)	12 (18.5)
	Prefer not to say	3 (2.4)	2 (3.2)	1 (1.5)
Employment status (n=127)	Full time	47 (37.0)	20 (32.3)	27 (41.5)
	Part time	28 (22.0)	13 (21.0)	15 (23.1)
	Unemployed	11 (8.7)	6 (9.7)	5 (7.7)
	Retired	39 (30.7)	23 (37.1)	16 (24.6)
	Prefer not to say	2 (1.6)	0 (0.0)	2 (3.1)
Treatment stage (n=130)	New patient, presurgical	45 (34.6)	23 (35.9)	22 (33.3)
	New patient, presurgical, neoadjuvant chemo	23 (17.7)	8 (12.5)	15 (22.7)
	New patient, postsurgical, adjuvant therapy underway	37 (28.5)	19 (29.7)	18 (27.3)
	Under follow-up, disease-free, maintenance treatment only	9 (6.9)	4 (6.3)	5 (7.6)
	Metastatic	16 (12.3)	10 (15.6)	6 (9.1)

Open in a separate window

GCSE, General Certificate of Secondary Education; NVQ, National Vocational Qualification.

Evaluation of the BRCA-DIRECT digital pathway

Uptake of BRCA-testing

Five (5/130) women withdrew from the study prior to receiving their pretest information. Of those who received pretest information, 123/125 (98.4%) consented to BRCA-testing; the two who withdrew were from the telephone (partially digital) arm (figure 2, online supplemental figure 3). See online supplemental table 2 for detail on withdrawals.

Hotline usage

The BRCA-DIRECT hotline was used overall by 24.7% of patients. Five ‘genetics specialist’ calls were made by patients (3.8% of all patients), seeking more information about what the results meant for the individual and their family members. Additional calls (63) were ‘administrative’ calls from 31 patients, with the majority relating to timing of results (n=35) and technical aspects of accessing the platform (n=18) (table 2).

Table 2

BRCA-DIRECT telephone hotline usage

Hotline calls recorded (n)				68
Average (mean) call length (min)				4.5
Patients using the hotline (n)				36
Proportion of participants using the hotline (%)				24.7
Study stage of hotline call			N	%
Prior to study consent			2	2.9
After study consent, up to receiving pretest information			22	32.4
After receiving pretest information, before completing genetic test consent			0	0.0
Awaiting results			37	54.4
After results/follow-up			7	10.3
Type of call			N	%
Genetics specialist		Pretest information	0	0.0
		Results	5	7.9
Administrative	Technical support	Sample provision	3	4.8
		Platform access or digital elements	15	23.8
	Process queries	Pathway/Study-specific	5	7.9
		Timing of results	35	55.6

Open in a separate window

Results turnaround

Samples were sequenced for all patients who completed their genetic test consent. Of the 123 results, there were 3 pathogenic variants, 0 likely pathogenic variants, 0 ‘hot’ VUS (5 evidence points) and 120 negative results. The overall median (IQR) time for testing of samples (from genetic-test consent to availability of results) was 27.6 (22.4–33.5) days, with similar turnaround times for patients in the fully digital (26.0 (20.6–33.2) days) and partially digital arms (28.6 (22.6–34.5) days) (online supplemental table 3).

Patient satisfaction with the pathway

Patient-reported satisfaction and perceived convenience for (a) pretest information delivery and (b) delivery of results was overall high with 86% of responses being ≥4 (5=most convenient/satisfied); see figure 3 for comparison between digital and telephone pretest information. Seven per cent of patients reported seeking assistance with accessing the BRCA-DIRECT platform from clinical/study staff (2.0% (2/100)) or friends/family members (5.0% (5/100)), and 13% sought assistance with providing a saliva sample, 12.0% (12/100) from clinical/study staff and 1.0% (1/100) from friends/family members (online supplemental table 4).

Open in a separate window

Figure 3

Patient-reported satisfaction and convenience. Patient Satisfaction Survey (PSS) completed by 100/130. Patient-reported convenience (1 not convenient–5 very convenient) and satisfaction (1 not satisfied–5 very satisfied). Digital (green) or telephone appointment (blue) delivery of pretest information (as per 1:1 randomisation) or results (as per random pre-allocation to telephone appointment for 2.5% of patients, plus for all patients receiving a variant of uncertain significance (VUS) or positive (pathogenic) result regardless of pre-allocation).

Patients accessed the BRCA-DIRECT digital platform by a smartphone alone (45.0%), desktop computer/laptop alone (24.0%), tablet alone (8.0%) or from some combination of devices (23.0%) (online supplemental table 4). Patient interviews revealed that reminder notifications were useful (all scored either 4 or 5, with 5 being very useful and 1 being not useful). Of those who received both SMS and email reminders, there was an equal balance preferring SMS compared with email notifications, with patients noting that SMS notifications acted ‘as a reminder’ and the emails enabled an easy link to complete questionnaires via a computer or laptop.

Healthcare professional satisfaction with the pathway

Eleven HCPs responded to the survey (18.2% clinical nurses; 18.2% consultant breast oncologists; 36.4% consultant breast surgeons; 27.3% other). The majority agreed (to varying extents) that all aspects of the BRCA-DIRECT digital pathway were equivalent (or superior) to standard-of-care, with the exception of end-to-end time-to-results (27.3% disagreed, 45.5% agreed, 27.3% neither agreed nor disagreed) (online supplemental table 5). Overall, 72.7% perceived the benefits of the pathway to outweigh the challenges of the pathway (online supplemental table 6) and 80.0% believed that the BRCA-DIRECT pathway is ready to be implemented in the NHS.

Discussion

We have presented data from our pilot of the BRCA-DIRECT pathway in the first 130 unselected patients with breast cancer from mainstream oncology services in 3 NHS hospitals, of whom half had the fully digital pathway (digital pretest information) and half had the partially digital pathway (telephone consultation pretest information). Considering the fully digital BRCA-direct pathway: uptake of BRCA-testing was high (60/64, 93.8%, with all withdrawals being prior to pretest information), as were ratings for perceived convenience and satisfaction for how they received pretest information and results (42/49 (87.8%) scoring as 4–5/5). Preliminary data regarding delivery of pretest information showed similar patient knowledge score, anxiety or satisfaction scores for the digital delivery and telephone genetic counselling.

As expected from general internet-access patterns, the BRCA-DIRECT platform was mainly accessed via a smartphone. However, a mixture of devices were used, demonstrating the importance of optimising the digital platform across different devices. Usability of the BRCA-DIRECT digital pathway was demonstrated to be high, with low numbers of patients requiring support, as indicated by both patient feedback (5.0% stated they sought technical support from another person) and analysis of calls placed to the hotline (23.8% of patients made a hotline call for administrative support regarding the platform). Notably, only 3.8% of patients accessed the hotline for expert genetics support, with all of these relating to results rather than pretest information. Similar hotline usage patterns have been reported by Gaba et al ³⁷ in unselected population-based personalised ovarian cancer risk assessment.

The majority of patient hotline calls placed were administrative and related to availability of results (35/68 (55.9%)). The time-to-results (median (IQR) time from receipt of sample (and study consent) to return of results) was 38.4 days (31.3–48.8) and testing turnaround time (time from genetic test consent to results available) was 27.6 days (22.4–33.5), reflecting the communicated turnaround time estimation of 3–4 weeks. However, the upper quartile of testing turnaround times experienced significant delays, reflective of the impact over this period of COVID-19-related supply chain issues for reagents and staffing shortages. Permissive estimates of turnaround time and provision on the digital platform of clear and accurate timeframes for turnaround of results is clearly critical for successful implementation and scaling of a fully digital pathway.

Five patients, out of 123, either failed to confirm that their results were received digitally or failed to book an appointment following notification of the results being available. Possible explanations included return of results coinciding with in-patient or treatment activity or patient demise. An alert was placed to the respective oncology professional, ensuring diversion to clinician-directed return of results. Such deviations illustrate the importance of integration of the digital pathway within oncology care delivery, ensuring both clinician awareness regarding patient progression with genetic testing and that results have been returned.

Acknowledgments

Footnotes

References