Late last week, AASLD published new practice guidance on the clinical assessment and management of NAFLD. The Surfers convene with Ken Cusi, who contributed to the previous iteration published in 2018, to explore its key features and implications. The updated document reflects the many advances pertinent to any practitioner caring for patients with NAFLD. This conversation introduces ideas around how the guidance describes use of biomarkers and NITs to either exclude advanced disease or identify those with a high probability of fibrosis.

Roger Green begins by recalling how the US addressed cholesterol 35 years ago. Reference labs were asked to change what was considered the normal value of cholesterol from two standard deviations above the mean down to a fixed number. This anecdote draws a parallel to the duality facing ALT. He suggests a lower high standard of 30U/L rather than 40U/L could be used, but recognizes that it might not be of benefit to emphasize ALT if patients with normal liver function also have NAFLD. He asks the panelists for their ideas around how we balance these issues to bring ALT into perspective. Jörn Schattenberg notes that typically the upper limit of normal is defined based on the reference population of where the test is evaluated. For example, his hospital in Mainz uses 50U/L on the basis of their reference population. Jörn explains that if this number is changed, an influx of abnormal tests will change the dynamics and economics of patients referred for suspected liver disease. On the other hand, reevaluating cut off points is relevant because patients with normal ALTs are far less likely to be investigated in terms of liver disease. His main takeaway: the guidance demonstrates that the advanced fibrosis disease stage would greatly benefit from the utility of FIB-4. Using this test at a large scale is sufficient and ultimately something to build on. Ken adds that the guidance has aimed to improve the sensitivity of FIB-4 in regards to false positives. He looks forward to the near future of increased precision where the right patients are appropriately referred to clinicians.

Roger notes that as a guidance document in the context of primary care, having three to four statements around ALT might be slightly confusing. He refocuses on the question of whether reference labs have a role in creating clarity around this issue. From a European perspective, Jörn suggests education is needed to drive the question from the physician to the lab. Ken points out that the US electronic medical record system should be able to integrate notifications around cut offs. Louise Campbell asserts the need for AST and ALT to be universally included within blood profiles. At the moment, this is not standard in many countries outside of the US. She harkens back to the potential AI has in the future of automated data analysis, but the essential first step is to ensure tests are able to be computed. In the final comment of this conversation, Ken reiterates the importance of recognizing and driving the message of harnessing what’s readily available today to positively influence the natural history of patient disease.