Relevant factors for policy concerning comprehensive genomic profiling in oncology: stakeholder perspectives

Abstract

Background

Comprehensive genomic profiling (CGP) can identify targets beyond standard of care, potentially revolutionizing personalized cancer management. However, conducting well designed studies in this rapidly evolving field is complex and demands time and investments. Consequently, the total added value of CGP remains uncertain. Clinical benefit and costs often are driving factors in coverage decisions. Recently, five additional factors were identified in the literature that can influence the choice for targeted profiling vs. CGP, specifically: “feasibility”, “test journey patient/physician”, “wider implications of diagnostic results”, “organization of laboratories”, and “scientific spillover”. The objective of the current study is to examine the role and importance assigned to these five additional factors for a comprehensive technology assessment by different stakeholders.

Methods

Purposive sampling was used to identify respondents from 4 stakeholder groups (i.e., medical specialists, molecular specialists, patient representatives, and policymakers) from different regions and hospital types (academic vs. non-academic) in the Netherlands. In semi-structured interviews, respondents scored the importance to decision-making of the five factors on a 0 (not important) to 5 (essential) scale. Reasoning behind the scores were elicited using open-ended follow-up questions. Transcripts were independently double-coded by two researchers using thematic analysis.

Results

Nineteen stakeholders (100% response rate; medical specialists (n = 7), molecular specialists (n = 7), patient representatives (n = 2), and policymakers (n = 3)) were interviewed. We observed differences between stakeholders in the relative importance assigned to the factors (range of median importance scores: 2–5). Overall, “wider implications of diagnostic results”, primarily CGP’s potential to identify additional treatment options, was deemed the most important factor alongside clinical benefit and costs in decision-making about CGP (median range: 3–5). While the “organization of laboratories” was considered less important (median range: 3–4), opposing arguments and preferences regarding the organization of laboratories were identified, with participants from academic centers preferring a centralized approach whilst non-academics preferred a decentralized approach.

Conclusions

Stakeholders deemed “wider implications of diagnostic results”, “feasibility”, and “test journey” the most important considerations for decision-making about targeted profiling vs. CGP alongside clinical benefit and costs. For policy decision-making, it is important to understand the arguments behind the heterogeneous opinions, often related to the setting they originate from.

Background

Molecular diagnostics is a cornerstone of precision oncology and widely used in advanced cancer management [1, 2]. Genomic profiling is applied to map genomic characteristics of tumors and subsequently identify patients eligible for targeted treatments or immunotherapies. In the last decade, there has been a substantial increase in approved targeted treatments and an expansion of indication areas for the use of immunotherapies. These developments can have considerable impact on the survival of eligible patients [3,4,5,6]. The molecular diagnostic pathway of patients can be highly heterogeneous, given the variety of molecular diagnostic techniques available in the hospital setting [7]. These techniques consist of single-gene sequencing methods (e.g., immunohistochemistry (IHC) or Fluorescence in Situ Hybridization (FISH)), targeted next-generation sequencing (NGS) panels and comprehensive genomic profiling (CGP, defined as NGS panels ≥ 200 genes; e.g., whole genome sequencing (WGS)). CGP can be applied in a tumor-agnostic fashion, independently from tumor type and stage. It covers genomic alteration beyond standard of care targets and can identify complex mutational signatures, such as tumor mutational burden [8].

In most countries, widespread reimbursement of CGP is limited. Consequently, access to CGP is often restricted to early access programs [9]. Positive reimbursement decisions by national healthcare institutes are needed to further enhance the implementation of CGP. However, clinical evidence of companion diagnostics is often immature at the time of appraisal for reimbursement [10]. Conventional randomized controlled trials are not always well-suited for the evaluation of new technologies, especially if like CGP, the technology undergoes rapid developments [11]. Additionally, compared to treatments, there often is a lower research budget available for diagnostic technologies, and thus, there is limited data on the desired outcomes available for decision-making. Meanwhile, a growing number of observational studies on CGP are published, which can spark enthusiasm in the clinic and among patients [12,13,14]. This growing enthusiasm places decision-makers in the position of having to determine whether to approve reimbursement for CGP, while evidence generation demonstrating the clinical benefit is still ongoing.

The potential value of innovative technologies might extend beyond clinical benefit, making it more challenging for decision-makers to determine how to comprehensively evaluate such technologies. Beyond the selection of on-label treatments and costs, we identified five factors through a literature review and expert consultation that might play a role in the choice for CGP: “feasibility” (adopting tests in the health care system), “test journey of patients and physicians” (pathway from requesting tests until reporting of results), “wider implications of diagnostic results” (impact of tests besides on-label treatments), “organization of laboratories” (organization of tests and access to tests), and “scientific spillover” (learnings through testing). Our review showed that for the majority of these factors, quantitative evidence was scarce [15]. There are multiple stakeholders with a stake in CGP reimbursement (e.g., medical specialists, molecular specialists, patient representatives, and policymakers) and it is unclear whether and to what extent the relative importance assigned to these factors differs between stakeholders. A qualitative evaluation of the relative importance assigned to these five factors by relevant stakeholders and their underlying considerations is lacking, but could help to further elucidate the role these factors could play in reimbursement decision-making for CGP. Therefore, the objective of this study is twofold, namely to: (1) assess the importance assigned to the five factors alongside clinical benefit and costs that could be relevant for decision-making about CGP by different stakeholders, and (2) get insights into stakeholders’ reasoning behind their relative importance score.

Methods

A qualitative study was conducted, using semi-structured interviews for the data collection.

Procedures and measures

Development of the interview protocol

A semi-structured interview guide was developed that consisted of three parts [16]. Part one consisted of questions to collect data on respondents’ characteristics (i.e., specialty, type of hospital they worked at, experience with molecular diagnostics, and research activities regarding (comprehensive) genomic profiling). In the second part of the interview, we asked respondents to share their views on the main advantages and disadvantages of CGP in general. In the third part of the interview, we went through a list of factors that, besides clinical benefit and costs, might be relevant in reimbursement decision-making for CGP. The identification of these factors is described in detail elsewhere. Briefly, factors were identified using a scoping review and confirmed in consultation sessions with experts. In this study, definitions of each of the factors were constructed according to the Good Practices for Multi-Criteria Decision Analysis to ensure that the factors are complete, non-redundant, do not overlap and are preference independent [17]. The list with factors is presented in Table 1. Respondents were asked to quantify the relative importance of each factor and to share the reasoning behind their score.

A few days prior to the interview, the list with factors was shared with each respondent. To understand the importance assigned to these factors from different stakeholders’ perspectives, respondents were asked to rank the importance of each factor in decision-making about whether to opt for targeted profiling or CGP on a Likert scale from 0 (not important) to 5 (essential) in a spider plot. Respondents did not explicitly conduct trade-offs between different factors. However, by ranking all factors simultaneously, they did consider the importance of factors relative to each other. Open-ended follow-up questions were asked to elicit the reasons for the relative importance score. Finally, respondents’ views regarding the relevance of the five factors compared to clinical benefit and costs was explored. The interview guide is listed in Supplementary 1.

Table 1 List with factors, alongside clinical benefit and costs, that might be relevant in reimbursement decision-making for comprehensive genomic profiling. WGS whole genome sequencing

The interview guide was pilot-tested with stakeholders (n = 5) from varying backgrounds (i.e., a pathologist, a clinical molecular biologist in the pathology, an oncologist, a clinical geneticist, and a hospital director). These pilot interviews were included in the final data analysis, as the adjustments to the interview guide were minimal (minor refinement to question phrasing). The study proposal was approved by the Antoni van Leeuwenhoek Internal Review Board (IRBd23-111) and all participants provided consent for the use of the interview data for the purposes of this project.

Recruitment strategy

Purposive sampling was used to select respondents from four stakeholder groups: medical specialists, molecular specialists, patient representatives and policymakers. The sampling procedure aimed to identify respondents from different regions and working at different types of hospitals (academic and non-academic) in the Netherlands. Given the highly specialist nature of the subject of the interviews, having knowledge/clinical experience with molecular diagnostic techniques in the context of oncology was an inclusion criteria. Participants were identified through the project teams’ network and invited to participate via email. The network primarily consisted of individuals involved in, or connected through, a national Dutch project aimed at improving the quality, affordability and accessibility of molecular diagnostics. The interviews were held from September to December 2022 and lasted between 30 and 60 min per interview.

Data analysis

Interviews were audio recorded with participants’ consent. Subsequently, they were transcribed verbatim with digital software (Condens) and the accuracy of transcripts was checked by LS. Transcripts were analyzed using a combination of deductive and inductive thematic analysis [18]. The focus of the analysis was to extract the reasoning to (not) test with CGP, clustering the reasons by the predefined five factors. If any reasons emerged that did not align with the predefined factors, they were assigned to a newly created category. The first five transcripts were used to develop a codebook, which was used to analyze the other interviews. New categories were added when encountered.

The transcripts were double-coded by LS and EE independently. Differences in coding were resolved through consensus, and if no consensus could be reached, a third reviewer was consulted (VR). Interviews were analyzed in batches of three to five interviews checking for newly identified topics that were to be examined in more detail in further interviews. Additionally, we assessed whether theoretical saturation had been reached for each stakeholder group separately. Descriptive statistics were used to summarize participant characteristics and to summarize the importance scores of the Likert scale questions for the five factors.

Results

In total, 19 stakeholders were interviewed from seven different regions in the Netherlands. The stakeholder group ‘molecular specialists’ consisted of three pathologists and four respondents who were either a molecular biologist in pathology or a laboratory specialist in clinical genetics (Table 2). The ‘medical specialists’ group consisted of four medical oncologists, one pulmonologist and two clinical geneticists. Three policymakers (a hospital director, insurer and policy advisor) and two patient representatives were interviewed. Of the participating molecular and medical specialists, 57% worked in the academic setting and 86% had ≥ 5 years of experience with molecular diagnostics.

Table 2 Respondent characteristics.^a only described for medical and molecular specialists, not for patient representatives and policymakers. CMBP clinical molecular biologists in pathology, LSCG laboratory specialist clinical genetics

Importance of factors from different stakeholder perspectives

Figure 1 provides an overview of the median and range of the importance scores assigned to the five factors by the different stakeholder groups, assessed using Likert scales. Overall, the “wider implications of diagnostic results” was scored as most important (median range 3–5), followed by “test journey” (median range 4–5) and “feasibility” (median range 3–5), while the “organization of laboratories” (median range 3–4) and “scientific spillover” (median range 2–5) were rated as the least important. The most marked differences in the importance score given by stakeholder groups were present in the “wider implications of diagnostic results” and “scientific spillover”. Respondents working in the academic setting attached a higher importance to “scientific spillover” than those working in non-academic settings. Specifically non-academic medical specialists thought the “wider implications of diagnostic results” of lower importance compared to the other stakeholder groups. The individual importance scores assigned by each respondent are listed in Supplementary 2.

Fig. 1

Stakeholder importance of the five factors. Median (presented by the circle) and range of the scored importance (0 = not important, 5 = essential) of the role of five factors in the choice for targeted profiling or CGP, assessed using Likert scales. Results are shown for patient representatives, academic and non-academic medical and molecular specialists. The results from patient representatives lack a range, as the two patient representatives were interviewed simultanously

Reasoning behind importance valuation

Feasibility

Most respondents did not consider feasibility a deciding factor for decision-making about CGP coverage. Although challenges in broad clinical implementation still exist, these were deemed surmountable if CGP was found to be of added value, and thus, not a deciding factor. Respondents anticipated these limitations to be addressed once the therapeutic advantages of CGP were more clearly established. Collection of fresh frozen (FF) biopsies (specifically needed for WGS) is an example of a surmountable barrier to widespread CGP implementation. Participants working in non-academic settings, indicated that even though it would be logistically challenging in the non-academic setting, most considered it feasible. The complex interpretation of CGP and the acquisition of the required knowledge was a frequently mentioned challenge. Participants illustrated this point by providing various practical examples, such as experienced inconsistency in quality between molecular tumor boards (MTBs) at different centers, a lack of clearly defined targets complicating the clinical translation of results, or commercial testing suppliers offering false hope to patients based on ambiguous targets.

Test journey from the patient and the physician perspective

The “test journey” for targeted profiling and CGP differed between the academic and non-academic setting. Stakeholders thought the turnaround time (TAT, time from requesting the test and receiving the results) should be below two weeks, to ensure a timely start of the next treatment line. In the academic setting, the TAT of both targeted profiling and CGP was considered to be acceptable. However, in the non-academic setting, where CGP typically is not offered as part of standard diagnostics, the TAT increased from < 1 week for targeted profiling to 3–6 weeks when using CGP. Foreseeing the lengthier TAT, non-academic medical specialists indicated that they had to anticipate to request CGP early in the process.

Repeated biopsies were mentioned by patient representatives as a major cause of stress. This was especially seen in patients referred to academic hospitals for additional diagnostics who had previously undergone targeted profiling in a non-academic setting. To avoid additional biopsies, several academic and non-academic molecular specialists preferred other CGP methods over WGS. This allowed them to utilize previously obtained Formalin-Fixed Paraffin-Embedded (FFPE) tissue, potentially saving patients an additional biopsy procedure. Non-academic specialists reported repeated biopsies for additional diagnostics to be rare in their setting and predominantly occur in tumor types requiring extensive testing, like cancer of unknown primary (CUP).

Molecular specialists believed that having all results in a single report, as opposed to several reports for separate tests, may be advantageous for clinicians. However, medical specialists themselves considered this a minor advantage, due to the guideline-based stepwise utilization of molecular diagnostic results and stepwise deployment of treatments. Patient representatives also emphasized the value of having a complete picture as provided by CGP, so that patients know whether there are any potential CGP-guided treatment options available to them in a later stage of the disease.

Wider implications of diagnostic results

All respondents, irrespective of background, saw the added benefit of CGP compared to targeted profiling in the identification of additional treatment options. Respondents agreed that the benefit of identifying these additional treatment options mostly applied to patients in good physical condition, who had exhausted all or most standard treatment lines. Clinicians in non-academic centers indicated that in their experience very few of their patients would be eligible for experimental, off-label or tumor-agnostic treatments. Consequently, they did not believe it to be necessary to test all patients for these potential treatment options. They emphasized the experimental nature of off-label and trial-based treatments, while in the academic setting these were seen as valid treatment options that should be accessible to patients. For example, a clinical geneticist from the academic setting stated; “It is about finding the optimal treatment and it does not matter to me whether that treatment is experimental or off-label”. In line with the academic perspective, patient representatives emphasized the value for patients of the identification of treatment options beyond standard of care.

Most respondents agreed that the other sub-factors of the wider implications of diagnostic results, namely the refinement of a diagnosis based on molecular profiles and pathogenic viruses, as well as the identification of hereditary cancer predisposition genes, pharmacogenetic profiles or predictive markers indicating a negative response, were less likely to play an important role in the choice between targeted profiling vs. CGP. Overall, there was consensus among stakeholders that the revision of a diagnosis should not be driven by molecular diagnostic results alone. However, academic respondents highlighted the added diagnostic value of CGP in refining the diagnosis in specific tumor types (e.g., CUP and sarcoma). While some academic specialists indicated simultaneous sequencing of germline material (WGS and whole exome sequencing (WES)) can streamline referrals to clinical geneticists, clinical geneticists themselves stated that additional germline findings on top of standard germline diagnostics would most likely consist of variants of unknown significance, and thus, be of limited clinical value.

Pharmacogenetic profiles, which can provide information on expected drug responses, metabolism or risks for adverse events, are currently not frequently applied in clinical practice according to medical specialists. However, molecular specialists have stated their ambition to include these in molecular panels. Contrary to opinions regarding the current use of pharmacogenetic profiles, medical specialists indicated that negative response markers are already integrated in targeted NGS panels. Currently used negative response markers in clinical practice are mostly restricted to KRAS, NRAS and BRAF. Consequently, negative response markers were not cited as a rationale for testing with CGP.

Organization of laboratories

While the median importance score for the “organization of laboratories” in the choice between targeted profiling vs. CGP was similar for academic and non-academic specialists (Fig. 1), opinions on the ideal way of organizing molecular diagnostics differed considerably between academic and non-academic specialists. CGP can contribute to a more uniform organization of laboratories, as more patients would receive the same diagnostic test. Academic molecular specialists thought this to be a major advantage, as it enhances testing efficiency by processing large volumes and thereby decreases costs. Additionally, they thought that uniform platforms simplified data exchange between different hospitals by ensuring comparable results, enhancing the learning ability between centers. However, non-academic specialists emphasized that maintaining variability in molecular diagnostic platforms increases the resilience of the healthcare system to external challenges, like shortages of certain manufacturers, while also acknowledging that reproducibility of results must be ensured between different platforms.

Differences in patient access to molecular diagnostics between the academic and non-academic setting was expressed as a concern by most academic specialists. A lack of expertise, insufficient national reimbursement or local financial hospital arrangements for molecular diagnostics were mentioned as causes for these differences. A few respondents stated that implementing CGP for all metastasized patients may increase equal access, as they thought that the heterogeneity in standard of care diagnostics results in suboptimal testing for some patients. Non-academic respondents, on the other hand, were less concerned about variations in patient access to molecular diagnostics, believing that most centers ensured sufficient access to molecular diagnostics. They considered it sufficient to only test for standard of care targets, while often omitting the low-incidence standard of care tumor-agnostic targets. Academic specialists think that for the optimal management of patients, tumor-agnostic and promising experimental treatment options should also be tested for.

The future-proofness of CGP, which involves incorporating promising targets to ensure the test’s usability for the upcoming period, was considered highly beneficial by academic molecular specialists. A future-proof test would reduce efforts required to introduce and validate new platforms when new biomarkers are introduced, while also allowing for ongoing development by creating new applications for the existing platform. Non-academic molecular specialists did not think that future-proof platforms were a valid reason for the widespread implementation of CGP. They indicated that even though the development of molecular biomarkers advances rapidly, integrating them into standard of care takes considerably longer. Additionally, the limited lifespan of sequencing platforms (approximately five years) would allow for a timely inclusion of newly introduced standard of care biomarkers.

Scientific spillover

Overall, most respondents recognized the scientific value of CGP and the potential contribution it could make to increasing molecular knowledge in cancer. However, specialists from non-academic hospitals thought that scientific spillover from CGP should be classified as research, and is therefore, not a strong argument to support the implementation of CGP. Several non-academic specialists argued that research should be performed through studies and should not be embedded into individual patient care. Respondents from both the academic and non-academic setting indicated that research purposes cannot be a reason to adopt interventions which increase public healthcare spending. Further, respondents from various perspectives underscored the necessity of establishing an infrastructure for using CGP results for research, which is currently lacking in the Netherlands. With regard to this infrastructure, clarity regarding data ownership and the trustworthiness of involved parties were mentioned by many stakeholders as concerns, especially when commercial parties are involved.

Health policy stakeholders

We interviewed three policymakers about the five factors identified that could play a role in the choice for targeted profiling vs. CGP. In general they felt less confident to rank the factors and indicated that some factors could best be addressed by physicians and as a consequence no descriptive statistics could be performed for this stakeholder group. During the interviews, they raised concerns about the financial impact of CGP on the healthcare system. It was indicated that the growth of the volume and costs of targeted treatments, and developments regarding the efficient use of these treatments are expected to be two important aspects influencing the magnitude of the financial impact of CGP. In the Netherlands, the healthcare system is based on managed competition. Insurers are expected to negotiate with health providers about quantity, quality of care and price. To effectively organize molecular diagnostics, it was indicated that insurers intend to move the field towards centralized diagnostic testing, which is expected to enhance testing efficiency, ensure high-quality diagnostics and decrease costs of testing. From an insurer perspective, centralization was also thought to enhance the learning cycle and improve knowledge development. To stimulate regional collaboration and harmonization between laboratories, insurers introduced Service Level Agreements, which contain regional agreements between (academic) laboratories concerning the quality and type of molecular diagnostics techniques used.

Given the financial budgetary constraints faced by non-academic hospitals within the Dutch financing system, the widespread implementation of CGP appeared doubtful. This was especially due to the availability of less expensive alternative targeted NGS panels and the current evidence levels supporting CGP. A prerequisite for the successful implementation of CGP was seen in building up appropriate decentral knowledge levels. Since non-academic centers treat the majority of patients, diffusion of knowledge about CGP was deemed to be necessary to guarantee equitable access to CGP for large groups of patients.

Discussion

In this study, we qualitatively explored the importance assigned to and arguments behind five factors that alongside clinical benefit and costs could be relevant for CGP reimbursement decision-making. The five factors were “feasibility”, “test journey of patients and physicians”, “wider implications of diagnostic results”, “organization of laboratories” and “scientific spillover”. Our results suggest that stakeholders from the academic and non-academic setting attach varying degrees of importance to the role these factors play in decision-making about CGP. Both academic and non-academic medical and laboratory specialists generally highly valued “wider implications of diagnostic results”, “feasibility”, and “test journey of patients and physicians”. “Scientific spillover” was found to be less important, especially by non-academic specialists. Patient representatives thought the “wider implications of diagnostic results” and the “test journey of patients and physicians” to be most important. Health policy stakeholders thought that patient volume, efficient use of expensive treatments, and decentral implementation were the more important considerations in the developments of molecular diagnostics.

“Wider implications of diagnostic results” was deemed the most important factor in decision-making for CGP, in particular the possibility to identify experimental, off-label or tumor-agnostic treatment options for patients. Academic stakeholders strongly advocated for equal patient access to CGP for the identification of these treatment options, while non-academic stakeholders emphasized the experimental nature. Numerous studies have been conducted providing CGP to patients, but the impact on overall survival remains uncertain due to the largely observational nature of these studies [15]. Innovative umbrella basket trials, such as DRUP, IMPRESS, ProTarget, or MEGALiT, are exploring the effectiveness of the tumor-agnostic use of registered targeted treatments [19,20,21,22]. The development of more robust evidence demonstrating the impact of a tumor-agnostic approach might be needed to align stakeholders’ perspectives on this factor of CGP.

Obtaining compelling evidence to demonstrate patient benefit of identifying biomarker-guided phase I/II/III trials might be more challenging due to the experimental and uncertain nature of such treatments. Nevertheless, it is worth noting that biomarker-guided phase I trials have demonstrated better response rates and improved progression-free survival compared to non-biomarker-guided-phase I trials [23]. This potential benefit, coupled with the immaturity of evidence of experimental treatments, poses challenges for reimbursement decision-making. In addition, future benefits through knowledge building, is seldom a formal criterion in reimbursement decisions. Following the principles of evidence-based medicine, decision-making frameworks tend to exclusively include proven interventions, to prevent the introduction of ineffective healthcare in the basic health insurance coverage, and therefore, do not consider this effect of CGP [24].

Perceived differences in patient access to molecular diagnostics between academic and non-academic stakeholders could be attributed, at least partly, to their perception regarding whether it is necessary to test for experimental mutations. In the Netherlands, CGP is offered in several academic centers or large laboratories, but no national initiatives are in place providing nation-wide access. Consensus on the minimally required targets that need to be tested for, or when CGP should be deployed, is essential to ensure patient access to an appropriate level of diagnostics in all centers. In most European countries, access to CGP is still limited. However, several countries have launched national initiatives to build CGP infrastructures and data ecosystems. These initiatives aim to enhance access and facilitate data sharing, thereby optimizing the benefits of genomics and acknowledging the potential for scientific spillover [25,26,27,28]. Currently, the application of CGP in a tumor-agnostic setting is only reimbursed in Norway, specifically for patients who have exhausted all treatment options [19].

The widespread implementation of CGP is complicated by the heterogeneous views on the preferred method of organizing molecular diagnostics. Efficiency improvements that CGP could offer, can only be achieved through a degree of centralization. This might even be a prerequisite for the successful widespread implementation of CGP to attain sufficient reductions in costs and turnaround times through economies of scale [29]. A system change towards CGP would be accompanied by procedural and financial uncertainties in many organizations, especially non-academic organizations who are less familiar with CGP. Considering these aspects in light of the NASSS (non-adoption, abandonment, scale-up, spread, sustainability) implementation framework, it is apparent that the organizational aspects of CGP adds to the complexity of the implementation process [30]. Empirical literature on the organizational impact of CGP is scarce [15], and more research in this area could reduce uncertainties for adopting actors (laboratories and hospitals) and facilitate implementation. Especially countries who opted for a top-down approach for the implementation of centralized CGP (e.g., France and the United Kingdom) have the opportunity to address some of the uncertainties regarding the consequences of a centralized organization.

Several limitations should be taken into account when interpreting the results of this study. First, the total number of respondents per stakeholder group was limited, but we ensured theoretical saturation was achieved within each clinical respondent group throughout the process. It was particularly challenging to identify respondents with sufficient expertise among policymakers and patient representatives. While these interviews yielded interesting findings, they might not fully represent the diversity of opinions within these stakeholder groups. Because the interview topic is highly specialized and requires expertise, purposive sampling was used to select respondents. This method might have introduced some selection bias, although we did identify distinctive perspectives and deliberately searched for candidates with various backgrounds. Lastly, although this study’s focus on the Dutch healthcare setting may somewhat restrict its generalizability to other countries, it is worth noting that the majority of findings are related to CGP in general and can be applicable to an international context, since the examined factors are not country-specific. Strengths of this study include the rigorous process applied to identify the five factors, which established a solid foundation for a more in-depth qualitative exploration in the interviews. We also included various stakeholder types from different settings to capture opinions from all relevant perspectives. This study is unique, as to our knowledge, this is the first study extensively describing the various potential benefits of CGP from different stakeholder perspectives.

We recommend that further research should be focused on addressing the heterogeneous views between academic and non-academic specialists regarding clinical and organizational factors of CGP. These issues can be addressed, by gathering more robust evidence demonstrating the benefits of CGP in both clinical and organizational aspects. In view of the dynamic nature of technology development in molecular diagnostics, we consider consistent, transparent and explicit inclusion of especially the factors “wider implications of diagnostic results”, “feasibility”, “test journey” and “organization of laboratories”, alongside the clinical benefit and costs, essential in reimbursement decision-making procedures.

Conclusions

According to multiple stakeholder perspectives, five previously identified factors, alongside clinical benefit and costs, play a role in the policy decision for targeted profiling vs. CGP. Especially “wider implications of diagnostic results”, “test journey”, and “feasibility” were deemed to be important for decision-making. Academic specialists thought additional experimental, tumor-agnostic or off-label treatment options as possibilities to improve and extend patients’ lives the most important reason to test with CGP, whereas non-academics considered CGP mostly as research with limited value for current clinical practice. While organizational factors were deemed less important in the choice of targeted profiling vs. CGP, academic and non-academic specialists had opposing views regarding how the preferred organization looked like. These are important considerations to take into account by decision-makers, as reimbursement of CGP could affect the current organization of molecular diagnostics thoroughly. This study highlights the different perspectives of relevant stakeholders on five factors, alongside clinical benefit and costs, and informs decision-makers on relevant factors and underlying reasoning to take into account in reimbursement decision-making about CGP.