Assessing Companion Diagnostics (CDx) for Reimbursement Within Alberta: Pilot Study Using Health Technology Assessment of Programmed Death Ligand 1 (PD-L1) Testing in Non-Small Cell Lung Cancer (NSCLC)

Jennifer Pillay
The number of investigational and licensed indications for companion diagnostic (CDx)-drug pairs is increasing. National assessments of the linked drug to support provincial funding decisions do not specifically address the CDx. The project objectives were to: (1) review a proposed process for Alberta to undertake health technology assessments (HTAs) of CDx and the rationale for a pilot, (2) provide background on the pilot HTA topic and identify the information elements for assessing CDx, (3)
more » ... ssessing CDx, (3) conduct the pilot HTA, (4) assess the proposed process for feasibility (timeliness, resources, and efficiencies), and compare the HTA findings with assessments in other jurisdictions/countries. Methods: Documents were reviewed describing a previous jurisdictional scan, literature review, workshop with Australian, British and American representatives, and process for CDx assessments as proposed to a provincial cross-sectoral, multidisciplinary CDx working group. A pilot of the proposed process using programmed-death ligand-1 (PD-L1) testing for advanced non-small cell lung cancer (NSCLC) was initiated. Using available guidance and literature, information elements related to CDx-specific to clinical utility, clinical validity and analytical validity-were identified. Research questions for the pilot HTA were developed using input from a topic working group, on clinical use of, and the laboratory capacity for PD-L1 testing in Alberta. For research questions about the prognostic role, clinical utility, analytical validity, costeffectiveness, and patient perspectives of PD-L1 testing for advanced NSLC, a systematic review using standard methodology was conducted. Other research questions (i.e., budget impact, social and ethical considerations) were answered based on literature from the systematic review and input/data from the topic working group. The HTA focused on use of the 22C3 PharmDx assay for PD-L1 testing to determine eligibility for pembrolizumab treatment in NSCLC. A process iii evaluation was conducted, based on: (i) data collection on timing, skill requirements and resources for all HTA steps, (ii) assessment (in retrospect) on how and when the information provided in reports by the national body providing recommendations on the linked drug(s) may be incorporated and/or impact the results, (iii) comparison of the findings with those of other countries. Results: The proposed process was designed to coincide with the pan-Canadian review of the drug and focus on information elements specific to the CDx. Findings from the pilot revealed that very low quality evidence exists for clinical utility of PD-L1 testing at the thresholds currently used for eligibility in advanced NSCLC for treatment with pembolizumab. PD-L1 expression may not greatly impact cost-effectiveness of the drug, because the benefits and costs change in the same direction as PD-L1 expression changes. The analytical validity of the PharmDx assay is sub-optimal; important considerations are that quality assurance is essential for local laboratories to undertake, and that a single biopsy in patients with multifocal lung cancer, as well as tissue samples from early stage disease, may not accurately capture PD-L1 expression status as used for treatment decisions. The budget impact to Alberta in 2017 was estimated at $535,296 annually when using actual cases submitted for testing (approx. 1,600); this cost may not be required if only cases of newly diagnosed advanced NSCLC (approx.. 450-500 per year) are tested in the future. The pilot was completed using 5.4 full-time equivalent months of effort (across personnel having expertise in systematic reviews, information science, and statistics), and could be conducted over a 5-to 6-month period to align with the current national assessments of the linked drug. Information on the cost-effectiveness and patient experiences with the CDx from the initial and/or final reports from the national assessments will likely prove very valuable. Assessment for reimbursement of the CDx alone, without concurrent iv and similar considerations applied for decisions about funding the drug, may lead to differences in access to the drug and CDx. This could limit the available options for funding the CDx and may lend towards lack of credibility to both processes. Conclusions: The proposed CDx review process appears feasible and generates the type and level of information required to support decision-making in Alberta. Rationale for decisions should be made transparent to assist with comparisons between provinces and other countries. v
doi:10.7939/r3mk65q8x fatcat:7537b43xn5cy3ogvsipxyaugdu