JPEN Journal Club 88. Determining the hypothesis of a study

Metabolic dysfunction-associated steatohepatitis (MASH), previously known as nonalcoholic steatohepatitis (NASH), has become one of the major causes for liver transplantation.1 This has led hepatologists to adopt MASH as a focus of their research and clinical interests. In this regard, it should be appreciated that only a small proportion of patients with MASH actually develop end-stage liver disease. Although such patients do have an increased risk of dying, those deaths are far more likely to be due to cardiovascular or malignant processes.2 The progression of MASH to various mortal outcomes (including liver failure) takes many years, leading investigators to employ various surrogate outcomes (in particular, histological improvements in the inflammatory and fibrotic components of MASH) as measures of efficacy. The US Food and Drug Administration (FDA) has recognized this problem and is accepting certain surrogate histologic outcomes as evidence to justify approving a drug for MASH. These outcomes include resolution of the inflammatory process and improvement (or at least stabilization) of fibrosis.3 Inflammation is assessed with the nonalcoholic fatty liver disease (NAFLD) activity score. This score is the sum of three histologic features (lobular inflammation [0–3 points], steatosis [0–3 points], and hepatocellular ballooning [0–2 points]) and can range from 0 to 8 (with a higher score representing more unfavorable histology). Resolution is defined as a ballooning score of 0, a lobular inflammation score of 0 or 1, and a total score reduction of ≥2 points from the baseline. Fibrosis is assessed on a scale from F0 (no fibrosis) to F4 (cirrhosis). It should be kept in mind that these surrogate outcomes have never been validated (ie, it has not been shown that improving them also improves mortality or morbidity). Not surprisingly, a number of potential pharmaceutical agents are now being tested in trials to identify beneficial effects on the histology of MASH. One of these randomized clinical trials (RCTs), by Sanyal et al.,4 evaluated survodutide. The report of that trial will be the subject of our discussion in this issue of the JPEN Journal Club. Sanyal et al. began their paper by noting that the FDA has already conditionally approved one agent, resmetirom, a selective thyroid hormone receptor beta agonist, because it improved the rate of resolution of the inflammatory component of MASH and also improved fibrosis.5, 6 (Sanyal et al. seemed to downplay this latter effect because the absolute difference in the rates of improvement between the treated patients and the placebo recipients was only 10%–12%). Glucagon-like peptide-1 (GLP-1) receptor agonists may also be useful because those compounds induce weight loss; survodutide is one such agent. However, hepatocytes lack GLP-1 receptors. Survodutide may be a potentially better option with regard to receptor agonists because it also is a glucagon receptor agonist, and such receptors are on hepatocytes (ie, survodutide is a dual-receptor agonist). Thus, these investigators undertook an RCT to assess this drug. Eligible participants for this trial were adults with biopsy-confirmed MASH (defined as a NAFLD activity score ≥4) and noncirrhotic fibrosis (stages F1–F3). The participants also had to have stable body weight, a liver fat content ≥8% as measured by magnetic resonance, and liver stiffness (a measurement of fibrosis) >6.0 kPa as measured by elastography. The study patients could not be using significant amounts of alcohol, be taking medications that are associated with hepatic abnormalities, or have any other chronic liver disease. Eligible participants were randomized into one of four groups, survodutide at doses of 2.4, 4.8, or 6.0 mg or placebo administered subcutaneously on a weekly basis for 48 weeks. The primary outcome was histologic improvement of MASH (≥2-point reduction in the total NAFLD activity score associated with a ≥1-point decrease in either lobular inflammation or hepatocyte ballooning) and no worsening of fibrosis. Other end points were a ≥30% reduction in liver fat, fibrosis improvement by ≥1 stage, absolute change in the NAFLD activity score, body weight, various laboratory tests and questionnaires, and adverse events. The investigators calculated that 240 participants would be required, but the investigators used a somewhat complicated method because they were also looking for a dose-response effect. In fact, all of the statistics were based on the possible existence of such an effect. 72 to 74 individuals were randomized into each group. The average age was 51 years, 53% were women, and the mean body mass index (BMI) was 36. Thirty-six percent had type 2 diabetes. The four groups were well matched with regard to a number of demographic and clinical features. The primary outcome was achieved in 47%, 62%, 43%, and 14% of the 2.4, 4.8, 6.0 mg, and placebo groups, respectively; the treated groups all had significantly better achievement rates compared with that of the placebo recipients. No comparisons were made between the three groups, and no comment was made regarding the presence or absence of a dose-response effect. Improvements in fibrosis were observed in 34%, 36%, and 34% of the three treated groups, compared with 22% in the placebo recipients. Liver fat content decreases of ≥30% occurred in 63%, 67%, and 57% of the groups receiving the survodutide but in only 14% of the control group. Body weight fell by 10%–13% in the treated groups but by only 0.6% in those receiving the placebo. Larger improvements in liver enzyme levels were also seen in the survodutide recipients. Over 90% of all four groups reported some type(s) of adverse event(s); the rates of nausea, vomiting, and diarrhea in the survodutide recipients were 2 to 10 times higher than that in the placebo group. Twenty percent of the three treated groups stopped the trial because of adverse events (mainly gastrointestinal), compared with only 3% of those using placebo. Sanyal et al. concluded that survodutide was better than the placebo with regard to improving the histology of MASH without worsening fibrosis. They did recognize the problem of adverse gastrointestinal effects and speculated that the symptoms might be ameliorated by a slower introduction of the drug. The only limitation to the trial that they noted was the fact that the participants were mostly White, so they could not generalize their findings to other ethnic groups. There was a companion paper in the same issue of the New England Journal of Medicine that reported an RCT of tirzepatide, another dual-receptor agonist (GLP-1 and the glucose-dependent insulinotropic polypeptide).7 The study design was similar although the patients could only have stages F2 or F3 fibrosis on entry. The average age of the participants was 54 years, and their mean BMI was also 36. Fifty-seven percent were female, and 58% had diabetes. Tirzepatide had a similar effect. To be appreciated, the mean body weight loss in those three treatment groups ranged from 11% to 15% compared with <1% in the placebo recipients. As I have noted before,8 it is being assumed by many hepatologists that MASH is a primary liver disease because it can lead to end-stage liver disease in some individuals. However, it is not clear to me that that is really the case any more than claiming that hypertension is a primary renal disease because it can lead to end-stage renal failure in some individuals. An alternative way to look at this issue is to consider MASH to be a manifestation of one or more metabolic conditions (eg, diabetes, obesity, hyperlipidemia, etc). If that is the case, we need to focus treatment on that (those) underlying condition(s). (With regard to the analogy to hypertension, although we do treat the blood pressure, we have not sought agents that interact with receptors in the kidney). In the case of these alleged specific MASH agents, the hepatic histologic manifestations clearly improved in conjunction with significant weight loss. Although the interest of hepatologists is largely driven by the number of patients who need liver transplantation, it must be appreciated that the incidence of liver failure in patients with MASH/NASH is low; the large number of liver failure subjects is a reflection of the high prevalence of MASH in the population. Assuming that these surrogate outcomes are valid, this does not mean that we should not use these newer agents; rather, we should consider their use in comparison with the relative costs, efficacies, and side effects of other interventions for these metabolic conditions. As an aside, it did appear that the differences in the rate of improvement of fibrosis was equivalent to what was reported with resmetirom. To the credit of Sanyal et al., they did not try to suggest that survodutide improved fibrosis. On the other hand, in the protocol of this trial (provided as an online supplement), it was stated4: "No confirmatory hypothesis testing is planned to be performed. Where P values are given, these are to be interpreted as exploratory." Their conclusions about the effectiveness of survodutide seemed somewhat inconsistent with this plan. Both protocols (one from 2020, before the trial began, and a second dated July 23, 2023, about 5 months before the trial ended) in the supplemental material stated that the main trial objectives were to demonstrate a nonflat dose-response curve, evaluate the size of the treatment effect, and characterize the dose-response relationship. This fact, as well as the introductory comments to the paper, led me to choose this paper to discuss. It was one of the few papers that I have encountered in which it was not clear to me what the actual question of the investigators was. The Introduction suggested that the investigators were wondering if the use of a dual agonist was better than just using a GLP-1 agonist. The protocol indicated that the first objective was to look for a nonflat dose-response curve. Most of the content of the paper indicated that the trial was assessing the utility of survodutide usage in MASH. Although it did appear that survodutide was better than placebo for improving certain unvalidated surrogate outcomes, there were no data that would allow us to compare that agent to one that was only a GLP-1 receptor agonist. It did not appear that there was any suggestion of a dose-response effect, but Sanyal et al. did not provide any comment about that issue. This conundrum is a reminder to us all that, although we look for the objective of a study in the text of the paper, in the end we will best define what the question is in any study by determining what question the methods will allow us to answer. You may have other thoughts about this paper. If so, I would encourage you to discuss them with your colleagues or even send an email to me ([email protected]). I would also again remind you that ASPEN has created an information kit with guidelines for how chapters, hospitals, and universities can develop their own journal clubs (www.nutritioncare.org/journalclub). For the next installment, please read the following article: Quigley EMM, Markinson L, Stevenson A, Treasure FP, Lacy BE. Randomised clinical trial: efficacy and safety of the live biotherapeutic product MRx1234 in patients with irritable bowel syndrome. Aliment Pharmacol Ther. 2023;57(1):81-93. The author declares no conflict of interest.