We appreciate the interest of Alkhouri et al. in this study.
The stated goal was to assess the performance of three MRI-based quantitative biomarkers in identifying any NASH and at-risk NASH in a cohort of patients with known or suspected NAFLD. As illustrated in Figure 5, we assessed these biomarkers individually and in multiparametric combinations. The paper does not anywhere suggest that imaging data should be used in isolation and thus we disagree with the suggestion that reporting these important results may add confusion to the field.
Contrary to the suggestion of Alkhouri et al., our study cohort consisted of patients with and without NASH (Figure 2). We emphasize therefore that our results pertain to a NAFLD population, not an exclusively NASH population. Our definition of at-risk NASH aligns with current clinical trial criteria and pharmacologic therapy enrollment standards (1).
We appreciate the reference to a previous study of similar MRI-based biomarkers (PDFF, MRE-LSM, and “iron-corrected” T1 measurement) (2). However, while that study reported the diagnostic performance of a combination of PDFF+T1 in detecting at-risk NASH, it did not report the performance of a combination of PDFF+MRE-LSM in the same task. Our results address that important gap.
Finally, we note that T1 is a basic physical property with a well-understood definition. Alkhouri et al. acknowledge that there are several different methods for measuring hepatic T1. Our study employed a widely available MRI method for T1 measurement that has been shown to provide robust results without some of the corrections that needed with other methods. Processing occurs on the MRI scanner, so the results are available immediately after the acquisition is complete. Unlike the proprietary “cT1” product referenced by Alkhouri, it is not necessary to send MRI data to a third party and pay an additional fee to obtain the results. To our knowledge, there is little or no published evidence indicating that the performance of the proprietary cT1 biomarker substantially exceeds that of properly acquired native T1 measurements in head-to-head comparisons for any relevant diagnostic purpose in NAFLD patient cohorts. A recent study in pediatric patients with autoimmune liver disease showed that cT1 measurements correlate with standard measurements of liver T1, supporting the general equivalence of these metrics with appropriate cut-offs (3).
In conclusion, we believe that our findings fully support our primary conclusions regarding the diagnostic performance of PDFF and MRE-LSM for identifying “any NASH” and “at-risk NASH” respectively, alone or in combination with other biomarkers, in patients with NAFLD.
Conflicts of interest:
Alina M. Allen consults, advises, and received grants from Novo Nordisk. She received grants from Pfizer and Target. Meng Yin owns stock and holds intellectual property rights with Resoundant. The remaining author has no conflicts to report.
Reference
- 1.Harrison SA, Allen AM, Dubourg J, Noureddin M, Alkhouri N. Challenges and opportunities in NASH drug development. Nature medicine. 2023;29(3):562–73. [DOI] [PubMed] [Google Scholar]
- 2.Imajo K, Tetlow L, Dennis A, Shumbayawonda E, Mouchti S, Kendall TJ, et al. Quantitative multiparametric magnetic resonance imaging can aid non-alcoholic steatohepatitis diagnosis in a Japanese cohort. World journal of gastroenterology. 2021;27(7):609–23. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Dillman JR, Serai SD, Miethke AG, Singh R, Tkach JA, Trout AT. Comparison of liver T1 relaxation times without and with iron correction in pediatric autoimmune liver disease. Pediatric radiology. 2020;50(7):935–42. [DOI] [PubMed] [Google Scholar]