Nonalcoholic fatty liver disease (NAFLD) has become one of the most commonly diagnosed liver conditions, detectable in almost a third of adults in developed countries, with substantially higher prevalence figures if the study population is narrowed to adults with type 2 diabetes mellitus and/or obesity [1]. Depending on the population under study, 10–20 % of NAFLD patients have the potentially progressive form of NAFLD known as NASH or nonalcoholic steatohepatitis. It is the latter group that has been the target of numerous ongoing therapeutic studies in an effort to slow disease progression to cirrhosis, liver failure, and hepatocellular cancer.

Therapeutic studies of NASH have been hindered by a number of factors, including a strong placebo effect, in part attributable to voluntary lifestyle changes, and limitations of noninvasive endpoints which rely on surrogate markers of cell injury instead of liver biopsy [2]. In contrast, liver biopsy itself has had inherent limitations as an endpoint. Most problematic has been the assessment of a key histological finding in NASH—cellular ballooning. Although ballooning is likely an important marker related to disease pathogenesis [3], agreement among observers regarding this histological hallmark can be surprisingly infrequent, especially absent special staining techniques [4, 5]. Thus, a diagnosis of NAFLD solely based on isolated histological parameters or even composite scores such as the NAFLD Activity Score (NAS) is questionable. Alternatively, although perhaps still debatable, the absence of NAFLD or the presence of advanced fibrosis stages is arguably more easily agreed upon histological parameters [6]. Moreover, the fibrosis stage itself carries prognostic significance [7].

In this issue of Digestive Diseases and Sciences, Glass et al. [8] present their experience with the effects of weight loss on fibrosis stage in a well-defined cohort of NASH patients followed with serial biopsy over a mean of approximately 5 years. Of 45 patients studied with a mean initial fibrosis stage of 1.96, 12 underwent bariatric surgery with Roux-en-Y gastric bypass with the majority experiencing significant weight loss. The most salient findings are nicely encapsulated in Fig. 1 where regression of fibrosis stage based on blinded biopsy review is most evident with the loss of ≥10 % total body weight (TBW), variable with 0–10 % TBW loss, with progressive increases more apparent among those experiencing weight gain. Most striking is regression from stage 4 to stage 3 and from stage 3 to stage 2 or even to stage 1. On the basis of multivariate analysis, fibrosis regression was predicted only by weight loss and not influenced by age, gender, glucose tolerance, ferritin level, or heterozygosity for α-1-antitrypsin type.

Fig. 1
figure 1

A number of variables which modulate disease progression or regression in NASH will likely influence the response to therapeutic interventions, including pharmacological therapy. A common thread linking these variables is their effect on fat metabolism, insulin sensitivity and resistance, and energy homeostasis

Although sampling error is a well-known problem with biopsy interpretation, taken in aggregate these results are convincing and support the emerging concept of the reversibility of even advanced stages of liver fibrosis [9]. Moreover, the results support the often challenging but fundamental tenets of NASH management to advise aggressive lifestyle changes with the encouragement of weight loss through dietary changes and exercise [10]. The broad application of bariatric surgery is more problematic due to concerns for costs, surgical eligibility, and perioperative risk with advanced fibrosis stages where postoperative complications can be substantial [11]. A recent study comparing the effects in NAFLD of Roux-en-Y gastric bypass (RYGB) versus the less-aggressive intervention of adjustable gastric banding (AGB) reported that RYGB was associated with greater sustained weight loss and more improvement in histological parameters of NASH [12]. Interestingly, however, fibrosis regression from stage 3 and 4 disease (bridging and cirrhosis) was about the same between the two interventions.

As an initial intervention, increasing physical activity and regular exercise is generally recommended in addition to dietary changes as a means to improve physical conditioning and calorie disposal. The most effective “dose” of exercise is uncertain, but one recent study using noninvasive surrogate markers of hepatocellular injury reported that moderate-to-vigorous exercise for at least 250 min/week was associated with the improvement that was, interestingly, independent of weight loss [13]. This level of exercise provides a target activity level that may, however, be difficult to achieve in NASH populations, as illustrated by the observation that physical deconditioning measured by bicycle ergometry was universally present in a cohort of NASH patients involved in a pharmacological study [14].

To date, no alternative interventions have been as effective in slowing or reversing fibrosis in NASH as weight loss of the magnitude reported in the current paper by Glass et al. Pioglitazone, a member of the thiazolidinedione (TZD) PPAR-γ agonists (peroxisome-proliferator-activated receptor “gamma”), is more effective in resolving histological NASH than is vitamin E or placebo and is associated with reduced fibrosis in meta-analyses [15, 16]. Peripheral weight gain, a potential and significant side effect of TZD therapy, is avoidable with efforts aimed at exercise and dietary changes [17]. Concerns over a risk of bladder cancer have proven to be unfounded [18]. Consideration of this medication is warranted either in selected patients in concert with endocrinological consultation or as part of further clinical trials. Other agents under current clinical study as potentially blunting or possibly reversing fibrosis related to NASH include simtuzumab (anti-LOXL2, [lysyl oxidase-like-2]), which is in advanced clinical trials (clinicaltrials.gov NCT01672866 and NCT01672879), and “GR-MD-02”, a galectin-3 inhibitor, which is another promising agent in early clinical studies (clinicaltrials.gov NCT01899859).

Other potential interventions that are commonly advised include changes in dietary composition to especially avoid beverages sweetened with rapidly absorbed high-fructose corn syrup which are particularly steatogenic in the liver [19]. Most but not all studies of ω-3 polyunsaturated fatty acid (PUFA) supplements including our own treatment trial have reported that these agents reduce liver fat without substantial changes in inflammatory activity or fibrosis [14, 20]. The variability in results, yet to be explained, could depend on whether steatosis is detected by global measures of liver fat content or solely by biopsy, or whether the level of steatosis varies among individuals. When we re-analyzed our data taking into account genetic variables, we found that the underlying genotype of fat metabolic pathways appears to influence the response to omega-3 supplements (manuscript submitted).

Given the variables that may modulate disease progression or regression (Fig. 1), it is likely that there will be individual variation in response to even the best treatment due to behavioral changes resulting from encounters with care providers and, likely, the genetic background of the individual. Factors include modifiable levels of physical activity (none to moderate or vigorous exercise), changes in total calorie intake and dietary composition interacting with lipoprotein metabolism, and key governors of hepatic steatosis such as PNPLA3 and possibly TM6SF2 [21]. Other environmental variables that have emerged as potentially significant include the gut microbiome and even cold exposure, which affect adipose tissue metabolism [22].

The study reported by Glass et al. confirms that advanced hepatic fibrosis in the setting of NASH is indeed reversible with significant loss of ≥10 % of body weight. Although most patients in the current report who achieved this endpoint had Roux-en-Y gastric bypass, the limitations of this intervention in terms of costs, eligibility, and potential complications mandate alternative approaches to affect a similar outcome. Ongoing work with weight loss agents appears promising, and even re-introduction of newer gastric balloons may offer alternatives [23, 24]. Future studies will likely require a measure of patience and discerning individualization to determine optimal therapy.