Assessing the cost & cost-effectiveness of next-generation sequencing for the diagnosis of hematological malignancies in children, adolescence & young-adults: A systematic literature review

Abstract Next-generation sequencing (NGS) is a cumulation of novel diagnostic modalities that can sequence entire parts of a cell in one sample. NGS is quicker and more accurate than standard-of-care (SoC) diagnostic methods, making it beneficial to cancer patients, particularly pediatric cancer, where a timely diagnosis is of the essence to improve survival outcomes, and because acquiring the adequate tissue samples needed for SoC modalities may be difficult, resulting in a compromised diagnosis. However, the integration of NGS into clinical settings has been limited due to the lack of existing data on its costs. We conducted a systematic literature review to measure the average total costs (ATC) and clinical outcomes of applying NGS modalities in comparison to SoC when diagnosing children, adolescents, and young-adults (AYA, 0-40 year) with cancer. We gathered English-language literature published beyond 2010. Study data had to be primary, focus on patients 0-40 years diagnosed with hematological malignancies via a next-generation sequencing modality, and to have included the ATC of their modality. Studies including older populations or those with other types of cancers were included if data regarding pediatric and AYA age group with hematological malignancies were provided. We gathered studies from MEDLINE, Embase, ResearchGate, and Google Scholar, and they were voted on Covidence. Nine studies were pulled to be reviewed, and we modelled their data on Microsoft Excel, Google Sheets, and RStudio. We modelled ATC per patient and ATC per sample, and we assessed clinically beneficial outcomes by determining the clinical validity, clinical impact, and potentially actionable findings (PAFs) derived from NGS. We also conducted a sensitivity analysis that calculated costs per patient for leukemia diagnoses after a 50% reduction in consumables prices, as well as predicted future costs of NGS for pediatric cancer diagnoses through a linearized exponential decay model. All nine studies report outcomes that were advantageous to patients and their treatments. Our findings show that on average, ATC per patient (across all NGS modalities) are approximately $6890 (in 2023 U.S. dollars), although separating the cost data from hematological versus non-hematological malignancies was not possible. The NGS modality with the lowest ATC per patient was RNA sequencing. Sensitivity analysis showed that ATC per patient for acute leukemia diagnoses can be lower than SoC modalities if the price of consumables decreases by 50%. The linearized exponential decay model constructed from the predictive data by Owens et al. demonstrates that ATC per patient could decrease by 82% from 2020 to 2030 (Adj R2=0.89, F=17.69, p=0.1486), although the high p-value indicates that more observations are needed to produce stronger results. A small number of studies and variations in reported units of measurement for clinical and cost data are major limitations in this review. Further, studies did not separate cost data for the different types of malignancies, nor for different age groups pediatric and AYA (0-40 year) versus adult (>40 year) patients. This may result in estimates for ATC to be over or underinflated. Although, at present, NGS costs are higher than the SoC methods for cancer diagnoses, available data predicts a reduction of ATC of NGS in the future. This systematic literature review highlights the importance of homogeneity among data collection methods to ensure comparability for future costing analyses. We recommend the development of uniform reporting guidelines for standardizing units of measurement for cost and clinical outcomes, and the application of different methods (such as the reduction in consumables prices) to reduce average total costs.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.