NASH Heterogeneity: Lessons From the Past and Future Directions

Key Takeaways

• Over the years, the focus of disease nomenclature has shifted from visual descriptions to genetic factors of underlying disease.
• Nonalcoholic steatohepatitis (NASH) is most likely heterogeneous in its underlying causes, with patients exhibiting differences in the predisposing genetic, epigenetic, and environmental factors.
• The heterogeneity of NASH indicates that patients may require individualized therapy based on the underlying cause.

For millennia, human diseases have been named by describing their appearance grossly (eg, erythema nodosum) or microscopically (eg, granulomatous hepatitis, lymphocytic colitis). This has served the useful purposes of clinical categorization and identification of causes, natural history, and treatments. However, as the tools for diagnosis and understanding of the underlying causes of disease have extended beyond the visual descriptions to a new era of molecular diagnostics, the shortcomings of visual descriptions have become apparent. Examples abound in oncology where the underlying genetic changes that govern cell survival and growth are being increasingly understood and leveraged therapeutically, with the nomenclature shifting from describing visual characteristics, such as small tumor cells vs large tumor cells, to disease definitions based on the underlying specific genetic abnormalities in genes that regulate cell growth, death, and responses to external stimuli (eg, estrogen receptor–positive, HER2-negative breast cancer).

Similar progress is being made in the field of hepatology. In the 1950s, when liver biopsies became used more frequently in the diagnostic evaluation of patients suspected of liver disease, an entity called “chronic active hepatitis” was created to describe liver biopsies with excessive inflammation, often concentrated in and around the portal tracts. By identifying a disease using a name, patients could be grouped into that histologic phenotype and treatments could be tested.

I had a budding interest in liver disease as a medical student at Yale, in no small part due to my sitting at the microscope with the legendary (which I really could not appreciate at the time) liver pathologist Gerald Klatskin. I recall being impressed by the varied histologic phenotypes of different liver diseases that he could quickly explain, and I also recall a then landmark study published in the New England Journal of Medicine four decades ago showed that corticosteroids can be beneficial in patients with chronic active hepatitis. However, steroids worked in some patients but failed in others. Now, decades later, we can appreciate in retrospect that some of those patients likely had autoimmune hepatitis and that they showed a great response to steroids. However, the treatment cohort likely included patients with various other forms of inflammatory liver diseases, including chronic hepatitis C, and such patients did not respond to steroids.

As the major causes of chronic active hepatitis became recognized, pathologists in the 1990s decided that, although the term aided in gathering patients with a similar liver biopsy phenotype for further study, it no longer served any useful purpose. They also decided that the term should be abandoned in favor of more precise terms that reflect the underlying cause, such as chronic hepatitis C, or some terms that may be more precise but are nonetheless descriptive and, thus, may still aggregate patients with poorly understood heterogeneity. Autoimmune hepatitis is certainly one such disease category that needs to be better understood at the molecular level to parse out the different genetic and exposure risks that lead to the histologic phenotype of liver inflammation with abundant plasma cells.

The Mayo pathologist, Jurgen Ludwig, in his seminal paper in 1980, coined the term “nonalcoholic steatohepatitis” (NASH) for a liver disease that is currently prevalent as an epidemic. The histologic phenotype of NASH is now well known after 4 decades of study and is characterized by the accumulation of excess triglyceride in lipid droplets of varying sizes, a mixed inflammatory infiltrate, and enlarged hepatocytes with rarified cytoplasm (observed after routine fixation), called ballooned hepatocytes, which may or may not contain aggregates of cytoskeletal proteins, called Mallory-Denk bodies. Fibrosis is the response to chronic inflammation in most tissues throughout the body. Fibrosis of varying degrees is often found in association with NASH and, in its most severe form, leads to death or need for liver transplantation in an alarming number of people.

Emerging data during the past decade have demonstrated that, like all other diseases that have been named virtually based on their appearance to the unaided eye or through the microscope, NASH is likely heterogeneous in its underlying causes. The implications of this pathogenic heterogeneity are substantial. It predicts that not all patients with the phenotype of NASH are likely to respond to any one treatment. It may also suggest that the noninvasive proteomic and lipidomic tests developed to identify patients with NASH may not be accurate in all patients if they reflect the underlying mechanism of disease rather than a consequence of the phenotype. The flip side of this conundrum is that such tests could be very valuable in subcategorizing patients with the phenotype of NASH and, hypothetically, could be useful in identifying patients with specific subtypes of NASH (or whatever we decide to call it in the future) that respond to some classes of drugs but not others.

Currently, the knowledge of potential subtypes of NASH is limited. Studies in humans and mice, based on liver and serum lipidomics, have shown that certain patients present with defects in methylation, an essential process for diverse processes such as the formation of phosphatidyl choline (an abundant membrane lipid) and regulation of gene expression. Other studies have separated patients with NASH into those primarily with metabolic disease vs those with high genetic risk scores, or impaired or intact very-low-density lipoprotein (VLDL) secretion. Whether these models of heterogeneity are different ways of looking at the same pathways or whether they overlap with each other needs further study.

The bottom line is that all patients with NASH are likely not the same in terms of the underlying predisposing genetic, epigenetic, and environmental contributors to this common histologic phenotype. It also means that when responders and nonresponders to specific therapies are identified in clinical trials, physicians are obligated to the patients in those trials and to all patients with NASH to use all tools available to understand why some respond and some do not. Only by doing so is it possible to engage in meaningful precision medicine for NASH. 

Your Thoughts?