Safety and Tolerability of Xanthohumol in Adults With Crohn's Disease: Results of a Triple‐Masked, Randomized, Placebo‐Controlled Phase 2 Trial

ABSTRACT

Xanthohumol (XN), a flavonoid from hops (Humulus lupulus), exhibits mucosal anti‐inflammatory, antioxidant, and microbiome‐modulating effects, making it a candidate therapeutic for inflammatory bowel disease. We assessed the safety and tolerability of XN in adults with Crohn's disease (CD) over 8 weeks. In this randomized, triple‐masked, placebo‐controlled phase 2 trial, 20 adults with unremitted CD received either 24 mg/day XN or placebo for 8 weeks. Primary outcomes included clinical laboratory toxicology parameters, vital signs and adverse events (AEs). Disease activity was screened and assessed using the Crohn's Disease Activity Index (CDAI). XN was well tolerated, with adherence exceeding 95%. Neither halting nor stopping criteria were met, and all laboratory elevations were minor and transient in both groups. There were no attributable serious AEs in either group, and all moderate AEs were short‐term and self‐resolving. Between‐group comparisons demonstrated differences for changes in BMI and gamma‐glutamyltransferase (GGT), favoring XN: BMI: −0.67 (0.0, 1.3), p = 0.04 and GGT: −4.8 U/L, 95% CI 0.8–8.8, p = 0.02, which may reflect beneficial effects on hepatic and metabolic health in CD. XN at 24 mg/day was safe and well tolerated in adults with active CD, supporting further research in inflammatory bowel disease.

Adults with active Crohn's disease (n = 20) received 24 mg/day of xanthohumol (XN), a bioactive compound derived from Humulus lupulus (hops), for 8 weeks. The study evaluated safety and tolerability through clinical laboratory measures, adverse events, vital signs, anthropometrics, health−related quality of life, and adherence. Stool, urine, and blood samples were collected to enable additional analyses of microbiome composition, short−chain fatty acids, calprotectin, XN metabolite profiles, and biomarkers of inflammation and gut permeability.

1. Introduction

Crohn's disease (CD) is a chronic, relapsing inflammatory bowel disease characterized by intestinal inflammation, dysregulated immune responses, and impaired barrier function. Standard therapies—including corticosteroids, immunomodulators, and biologics—while effective, carry risks of serious adverse events (AEs), including infection, hepatotoxicity, and malignancy [1, 2]. The therapeutic ceiling achievable by current therapies justifies the continued development of therapeutics for inflammatory bowel disease (IBD) [3], including safe, adjunctive therapies to modulate inflammation and support mucosal health. Microbiota‐ and diet/microbiota‐driven mechanisms in IBD provide a rationale for the development of therapeutics based on their impact on the microbiota and its metabolism [4, 5]. This includes evaluation of potential mechanisms of short‐chain fatty acid production, gut barrier function, anti‐inflammatory activity, and the potential to repair [6, 7, 8].

Xanthohumol (XN), a prenylated flavonoid derived from the female inflorescences of Humulus lupulus (hops), has demonstrated mucosal anti‐inflammatory, antioxidant, and microbiome‐modulating properties in preclinical and clinical studies [9, 10, 11, 12, 13]. XN exerts upstream regulatory effects on key inflammatory pathways, including inhibition of NF‐κB and activation of Nrf2‐mediated antioxidant responses. Additionally, XN interacts with the farnesoid X receptor (FXR), an important regulator of bile acid metabolism, intestinal permeability, and mucosal immunity [11]. Phase 1 clinical results [14] also demonstrated XN‐mediated changes in bile acid metabolites [10]. Specifically, XN modified bile acid profiles and reduced microbiota‐derived secondary bile acids in an enterotype‐dependent manner in healthy adults. These findings are noteworthy given the role of bile acids in intestinal inflammation and dysbiosis.

XN may also be a novel candidate therapeutic in IBD due to its effects on cardiometabolic inflammation, which may be caused by increased gut barrier permeability and resultant endotoxin/lipopolysaccharide‐induced increases in proprotein convertase subtilisin/kexin type 9 (PCSK‐9) [15, 16, 17]. XN and other polyphenolics have demonstrated anti‐PSCK‐9 activity [18, 19], which may be mediated by improvements in gut barrier function [20, 21]. Because cardiovascular event risk is increased in people with IBD [22, 23, 24], compounds that simultaneously reduce inflammation, increase functional activity, and reduce permeability of the gut would be promising therapeutics. XN and other polyphenolics hold particular potential for individuals with evidence of metabolic hepatic inflammation [25] due to demonstrated effects on reducing hepatic inflammation and vascular atherogenic activity.

The current phase 2 study [1] was designed to build upon our prior phase 1 trial that demonstrated safety and tolerability of 24 mg/day XN in healthy adults [14]. Here, we aimed to evaluate whether XN retains a favorable safety and tolerability profile in adults with clinically active CD while also exploring changes in mechanistic biomarkers such as the liver function test γ‐glutamyltransferase (GGT), an indicator of metabolic hepatic dysfunction and cardiovascular disease (CVD) risk [26, 27].

2. Experimental Section

2.1. Study Design

We conducted a randomized, triple‐masked, placebo‐controlled trial under IND#140626 from the U.S. Food & Drug Administration (FDA) as published [1] and registered on clinicaltrials.gov as protocol #NCT04590508. In brief, participants enrolled in the trial were randomized to take XN or placebo daily and attended clinical research visits every 2 weeks for 8 weeks. Urine, stool, and blood were collected for measurement of XN metabolites, microbiota sequencing and clinical labs and cytokines, respectively. The trial protocol and corresponding Data Safety Management Plan (DSMP) were approved by the FDA, the National Center for Complementary and Integrative Health (NCCIH), and the IRB at the National University of Natural Medicine. A standardized informed consent form was approved by the IRB and administered to all participants, including allocated time for questions; all participants provided informed consent prior to enrolment and randomization.

2.2. Participants

Adults, 21–70 years, with clinically active CD, defined as a Crohn's Disease Activity Index (CDAI) score > 150, were recruited by the Helfgott Research Institute (Portland, OR) from the region between Seattle, WA, to Salem, OR. Key reasons for participant exclusion included biologic therapy, high‐dose steroids, and/or use of polyphenolic, cannabinoid, or probiotic supplements. Detailed participation criteria are reported in our previously published protocol [1]. A total of 20 participants were randomized 1:1 to XN or placebo. Allocation concealment was accomplished by group assignment being contained in a signed and sealed envelope until the moment of randomization, at which time group assignment became known to the Study Coordinator and the participant.

2.3. Intervention

XN capsules contained 24 mg ≥ 99% pure XN, encapsulated with rice protein. The placebo contained only rice protein. Participants took one capsule orally with food daily. Adherence was assessed by capsule count at each clinical research visit. Capsule contents were measured for XN amounts every 2 months by HPLC with UV detection using the European Brewery Convention method 7.15: Xanthohumol in Hops and Hop Products by HPLC.

2.4. Outcome Measures

Primary outcome measures of this Phase 2 trial included self‐reported clinical AEs, vital signs, and clinical laboratory biomarkers. Laboratory measures of hematologic safety [red blood cell count (RBC), white blood cell count (WBC), platelet count (PLT), hemoglobin (Hgb), hematocrit (Hct)], liver function [aspartate aminotransferase (AST), alanine aminotransferase (ALT), gamma‐glutamyltransferase (GGT)], and renal function [estimated glomerular filtration rate (eGFR), blood urea nitrogen (BUN) and creatinine (Cr)] were the pre‐specified primary outcome measures for safety determination. Subjective measures of safety included self‐reported health‐related quality of life, specifically the PROMIS‐29 [28], the participant/physician‐assessed CDAI [29, 30], which includes self‐reported symptom frequency and also physical exam findings. The primary endpoint was pre‐specified as 8 weeks.

AEs were collected and graded based on an 8‐system, 90‐question standardized symptom interview with grading definitions based on NCI Common Terminology Criteria for Adverse Events v4.0. Because they impacted halting and termination decisions, AEs were tracked in a separate database and updated in near real time shortly after each participant's bi‐weekly research visits.

The protocol [1] specifically defined clinically significant changes in laboratory measures: “Elevations” of liver function tests were considered “significant” if they were: (1) abnormal per laboratory reference ranges, and (2) represented a 25% deviation from previously observed values. New onset anemia was considered “significant” if: (1) hemoglobin < 14 g/dL (140 g/L), hematocrit < 42% (< 0.42), or RBC < 4.5 million/µL (< 4.5 × 10¹²/L) (for men); or hemoglobin < 12 g/dL (120 g/L), hematocrit < 37% (< 0.37), or RBC < 4 million/µL (< 4 × 10 ¹²/L) (for women); and (2) values represented a 10% change or more from baseline values; and (3) values remained abnormal with repeat testing within 2 weeks. This threshold reduced reporting of new “borderline” abnormals for participants with “high normal” baseline values. For elevations of creatinine, “significant” was evaluated as greater than a 20% increase from baseline values, whereas reductions in eGFR were considered “significant” if greater than a 20% decrease from baseline values. A brief clinical interview was conducted and recorded for all laboratory abnormalities not present at baseline to provide context, for example, new onset anemia tied to increased diarrheal episodes, and so forth.

2.5. Statistical Analysis

The trial sample size (enrolment of up to n = 32 to retain an evaluable sample size n = 24) enabled powered (> 80%) measurement of a 25% difference in the proportion of participants who experienced an elevation in any pre‐specified clinical laboratory measures at a significance threshold alpha = 0.05. Between‐group (XN vs. placebo) comparisons of mean changes in each safety parameter from baseline to Week 8 were conducted using unpaired t‐tests. Due to low incidence, AEs are reported descriptively and were not compared statistically. Also, the threshold for statistical significance (i.e., alpha = 0.05) was not adjusted for multiple comparisons, such that interpretation of safety remained more conservative. This analytical plan was approved a priori by the FDA, NCCIH and the IRB.

3. Results

3.1. Participant Characteristics

The final cohort (n = 20) was 60% female, mean age of 38.6 ± 15.2 years. A CONSORT‐style enrolment and retention flowchart is included as Figure 1 below. Common reasons for exclusion included: time restraints, CD in remission, use of prohibited dietary supplements or nicotine, and not having a primary care provider. Baseline demographics are shown in Table 1 and were balanced between groups. Recruitment was suspended after 2.5 years of total active recruitment, including a final year of recruitment efforts without enrolling a participant.

FIGURE 1.

CONSORT enrolment and retention flowchart.

TABLE 1.

Baseline characteristics.

Characteristic	Placebo (N = 10) Mean (SD) or N (%)	Xanthohumol (N = 10) Mean (SD) or N (%)
Age (years)	36.20 (14.0)	41.00 (16.7)
Sex
Male	3 (30%)	5 (50%)
Female	7 (70%)	5 (50%)
Ethnicity
Non‐Hispanic/Latino	10 (100%)	7 (7%)
Hispanic/Latino	0 (0%)	3 (30%)
Race
White/Caucasian	9 (90%)	9 (90%)
Asian	0 (0%)	1 (10%)
Black/African American	1 (10%)	0 (0%)
Height (m)	1.71 (0.01)	1.68 (0.1)
Weight (kg)	86.94 (21.6)	74.51 (22.5)
BMI (kg/m2)	30.01 (8.0)	26.64 (8.1)
CDAI	220.7 (46.3)	211.0 (38.2)
PROMIS‐Physical Function	49.9 (8.1)	47.6 (7.2)
PROMIS‐Fatigue	55.4 (8.1)	55.4 (8.5)
PROMIS‐Pain Interference	55.3 (9.2)	56.3 (7.6)
PROMIS‐Pain Intensity	3.90 (1.9)	3.20 (1.9)
PROMIS‐Anxiety	54.9 (11.0)	50.0 (8.0)
PROMIS‐Depression	50.2 (9.7)	48.1 (8.5)
PROMIS‐Social Function	49.1 (10.0)	54.2 (6.8)
PROMIS‐Sleep	56.7 (2.4)	54.5 (5.3)

Note: CDAI max is 500 with < 150 considered clinical remission, PROMIS scores are t‐scores with mean 50.

Abbreviations: %, percentage; BMI, body mass index; kg, kilograms; kg/m², kilograms per meters squared; m, meters; N, number; SD, standard deviation.

3.2. XN Stability

Monthly throughout the recruitment period and trial duration, all randomly selected XN capsules from the trial production batch contained XN within 10% deviation from the nominal amount (24 mg), in accordance with USP guidelines.

3.3. Safety Outcomes

One participant withdrew from the XN group due to a self‐reported flare in their CD. Although attribution due to XN cannot be ruled out entirely. Notably, the participant did not have any laboratory abnormalities at the prior week's visit, though they did report moderate abdominal pain. The study team remained in contact with the participant following their withdrawal at Week 6. The participant was queried regarding improvement since discontinuing their intervention; however, the participant reported there had not been any improvement since stopping the intervention. Although not formally measured by a validated self‐report instrument, the participant subjectively reported an increased burden of stress due to family and work circumstances at the time of their withdrawal, confounding attribution. The frequencies of all self‐reported AEs by visit and grade are reported in Table 2.

TABLE 2.

New‐onset adverse events by group and grade.

	Xanthohumol (n = 9)				Placebo (n = 10)
Week n per grade of severity (1|2|3)	Week 2 (1|2|3)	Week 4 (1|2|3)	Week 6 (1|2|3)	Week 8 (1|2|3)	Week 2 (1|2|3)	Week 4 (1|2|3)	Week 6 (1|2|3)	Week 8 (1|2|3)
Nystagmus	1|0|0
Nasal congestion						1|0|0	1|0|0
Phlegm					1|0|0
Allergies	1|0|0	3|1|0	1|1|0	1|0|0
Abdominal pain			1|2|0	0|1|0	1|0|0		0|1|0	0|1|0
Decreased appetite	1|0|0	1|0|0	1|0|0	1|0|0		1|0|0		2|0|0
Increased appetite	2|0|0	1|0|0	1|0|0		1|0|0
Constipation	1|0|0	1|0|0		1|0|0		1|0|0	1|0|0
Diarrhea	1|1|0	0|1|0	1|1|0					0|1|0
Nausea			0|1|0		2|0|0		2|0|0	1|0|0
Emesis							1|0|0
Increased thirst	1|0|0
Decreased thirst			1|0|0				1|0|0
MSK pain						1|0|0
Headache	3|0|0	2|0|0	1|2|0		1|0|0	1|0|0	1|0|0
Restlessness
Anxiety		1|0|0			1|0|0	1|0|0		1|0|0
Depression								1|0|0
Difficulty concentrating				1|0|0
Irritability					2|0|0
Drowsiness	1|0|0				3|0|0
Lethargy				1|0|0	2|0|0	1|0|0
Excessive sleep	1|0|0				1|0|0	1|0|0
Insomnia	1|0|0
Fatigue	2|0|0		0|2|0		2|0|0	2|0|0	1|0|0
Dyspnea	0|1|0							1|0|0
Chest pain	1|0|0
Difficult/Labored breathing								0|0|1
Acne			1|0|0
Erythema		1|0|0			1|0|0
Pruritus	0|1|0	2|0|0	1|0|0	1|0|0	1|0|0
Local edema	1|0|0
Rash	0|0|1
Easy bruising					1|0|0	1|0|0
Increased sweating	1|0|0
Increased libido			1|0|0
Decreased urination					1|0|0
Difficult urination								0|1|0
Dizziness			1|0|0		1|0|0
Fever	1|0|0					0|0|1
Sore throat	1|0|0
Cough	0|0|1
Myalgias	1|0|0
Grade 1	22	12	11	6	22	11	8	7
Grade 2	3	2	9	1	0	0	1	3
Grade 3	2	0	0	0	0	1	0	1
Total	68				53

In the XN group, 68 total AEs were reported, including 51 “Grade 1/Mild,” 15 “Grade 2/Moderate,” and 2 “Grade 3/Severe.” Grade 3 AEs included one case of novel coronavirus infection and one outbreak of recurrent herpes zoster. Both cases were treated for cough and antiviral therapy, respectively, by their physicians.

In the placebo group, 54 total AEs were reported, including 48 “Grade 1/Mild,” 3 “Grade 2/Moderate,” 2 “Grade 3/Severe,” and 1 “Grade 4/Serious.” Grade 3 AEs were separate cases of novel coronavirus infection; Both participants remained in the study and completed final follow‐up assessments. The single Grade 4/Serious AE was hospitalization due to bilateral pulmonary embolism, which was managed with anticoagulant therapy and resolved. This isolated SAE was classified as “unexpected” but “not attributable” to the placebo.

Comparing self‐reported AEs between the XN and placebo group, the difference in “Grade 2/Moderate” AEs stands out as a potential concern, particularly at the Week 6 visit, that is, nine in the XN group versus one in the placebo group. Grade 2 AEs are defined as symptoms requiring the use of over‐the‐counter treatments and/or a single physician's visit, and negatively impacting activities of daily living (ADLs) for less than 48 h. In this case, the symptoms that were higher in the XN included allergies (1 vs. 0), abdominal pain (2 vs. 1), diarrhea (1 vs. 0), headache (2 vs. 0) and fatigue (2 vs. 0). Although attribution to XN cannot be entirely ruled out, it is important to note that one participant began experiencing a self‐reported flare, including abdominal pain, diarrhea and fatigue at Week 6, and ultimately withdrew at their Week 6 visit, and thus three of the “moderate” AEs were in a single participant. Similarly, allergy symptoms with corresponding headache were also reported by a single participant. Given these isolated and concurrent cases, the remaining AE reports were single reports of increased abdominal pain, headache, and fatigue. Also reassuring is that all moderate‐severity symptoms, except a single case of residual abdominal pain, resolved by Week 8.

Laboratory markers are detailed in Table 3 and revealed no significant abnormalities. GGT was significantly reduced in the XN group compared to placebo at Week 8 (mean Δ: −4.8 U/L, 95% CI 0.8–8.8, p = 0.02). No other significant changes in clinical laboratory outcome measures were observed.

TABLE 3.

Pre‐specified primary safety outcome measures: Mean change baseline to Week 8.

	Placebo (N = 9)		Xanthohumol (N = 10)		Mean Δ (95% CI) **
	Week 0 Mean (SD)	Week 8 Mean (SD)	Week 0 Mean (SD)	Week 8 Mean (SD) *	Baseline—Week 8	p a
Vitals/Anthropometrics
Systolic BP (mmHg)	124.80 (16.0)	120.50 (15.2)	126.44 (14.9)	125.67 (15.4)	3.52 (−13.9, 6.9)	0.48
Diastolic BP (mmHg)	73.40 (12.7)	73.90 (10.7)	74.56 (9.1)	75.44 (9.3)	0.39 (−6.88, 6.09)	0.90
Heart rate (bpm)	74.40 (11.4)	77.00 (5.4)	72.56 (8.3)	78.56 (8.4)	3.4 (−12.3, 5.5)	0.43
Temperature (°F)	98.37 (0.6)	98.01 (0.5)	98.38 (0.6)	98.30 (0.5)	0.28 (−0.8, 0.3)	0.29
BMI	30.01 (8.0)	30.40 (8.2)	27.30 (8.3)	27.02 (7.8)	−0.67 (0.0, 1.3)	0.04
Comprehensive chem panel
BUN (mg/dL) RR: 7–25 mg/dL	12.50 (4.0)	12.80 (3.0)	16.11 (6.9)	16.22 (8.0)	−0.19 (‐3.3, 3.6)	0.91
Creatinine (mg/dL) RR (M): 0.7–1.10 RR (F): 0.5–1.10  mg/dL	0.88 (0.2)	0.90 (0.2)	0.95 (0.30)	0.93 (0.2)	−0.03 (−0.1, 0.1)	0.53
eGFR  mL/min/1.73 m2 RR: > 60 mL/min/1.73 m2	95.20 (24.5)	93.50 (22.3)	93.11 (31.9)	95.22 (29.6)	3.81 (−14.5, 6.9)	0.46
Sodium (mmol/L) RR: 135–146 mmol/L	139.20 (2.0)	138.80 (1.6)	139.22 (1.1)	140.33 (2.6)	1.51 (−3.4, 0.4)	0.12
Potassium (mmol/L) RR: 3.5–5.3 mmol/L	4.24 (0.5)	4.17 (0.2)	4.39 (0.6)	4.30 (0.6)	−0.02 (−0.5, 0.6)	0.94
Chloride (mmol/L) RR: 98–110 mmol/L	104.00 (1.9)	103.90 (2.2)	104.00 (4.9)	103.67 (3.8)	−0.23 (−2.3, 2.7)	0.85
Calcium (mg/dL) RR: 8.6–10.3 mg/dL	9.36 (0.5)	9.35 (0.3)	9.48 (0.3)	9.70 (0.4)	0.23 (−0.6, 0.1)	0.17
Bilirubin (mg/dL) RR: 0.2–1.2 mg/dL	0.62 (0.2)	0.62 (0.3)	0.71 (0.4)	0.67 (0.3)	−0.04 (−0.1, 0.2)	0.63
Alkaline phosphatase   (U/L) RR: 40–115 U/L	52.20 (7.3)	54.50 (9.0)	74.22 (17.7)	77.44 (20.9)	0.92 (−8.4, 6.6)	0.80
AST (U/L) RR: 10–35 U/L	19.30 (9.9)	20.00 (10.9)	22.22 (8.4)	20.56 (6.3)	−2.37 (−4.6, 9.3)	0.48
ALT (U/L) RR: 9–46 U/L	19.00 (12.3)	20.20 (11.4)	27.78 (19.3)	21.11 (11.4)	−7.87 (−0.8, 16.6)	0.07
GGT (U/L) RR: 9–48 U/L	22.50 (18.9)	24.30 (21.6)	21.89 (12.7)	18.89 (9.0)	−4.8 (0.8, 8.8)	0.02
Complete blood count
WBC (K/uL) RR: 3.8–10.8 ×103/µL	7.64 (3.2)	7.14 (2.0)	6.82 (2.4)	7.02 (2.2)	0.70 (‐2.2, 0.8)	0.33
RBC (M/uL) RR (M): 4.20–5.80 × 106/µL RR (F): 3.8–5.1 × 106/µL	4.72 (0.3)	4.61 (0.2)	4.95 (0.5)	4.88 (0.3)	0.05 (−0.2, 0.1)	0.58
Hemoglobin (g/dL) RR(M): 13.2–17.1 g/dL RR (F): 11.7–15.5 g/dL	13.75 (1.5)	13.62 (1.3)	14.84 (1.5)	14.71 (1.4)	−0.00 (−0.5, 0.6)	0.99
Hematocrit RR (M): 38.5%–50% RR (F): 35%–45%	42.12 (2.9)	41.55 (2.6)	44.70 (4.1)	44.22 (3.9)	0.09 (−2.2, 2.0)	0.93
Platelets (K/uL) RR: 140–400 × 103/µL	298.70 (68.1)	308.30 (82.7)	278.00 (76.7)	308.78 (103.7)	21.18 (−68.8, 26.5)	0.36

^*One participant in the xanthohumol group had incomplete for Week 6 and Week 8; thus, n = 9 for Week 8 in xanthohumol group.

^**Mean Δ is reported as the difference in each parameter from Week 8 to Baseline in the XN group minus the difference in the placebo group.

^a p value calculated as mean change in value from baseline by unpaired, two‐sided t‐test between groups.

Abbreviations: °F, degrees in Fahrenheit; ALP, alkaline phosphatase; ALT, alanine aminotransferase; AST, aspartate aminotransferase; BP, blood pressure; bpm, beats per minute; BUN, blood urea nitrogen; CI, confidence interval; CRP, c‐reactive protein; EGFR, estimated glomerular filtration rate; F, female; fL, femtoliter; g/dL, grams per deciliter; GGT, gamma‐glutamyl transferase; K/uL, kilo units per microliter; M, male; M/uL, million per microliter; MCH, mean corpuscular hemoglobin; MCHC, mean corpuscular hemoglobin concentration; MCV, mean corpuscular volume; mg/, milligrams per liter; mg/dL, milligrams per deciliter; mmHg, millimeters of mercury; mmol/L, millimoles per liter; MPV, mean platelet volume; pg, picograms; RBC, red blood cell; RDW, red cell distribution width; RR, reference range;  SD, standard deviation; U/L, units per liter; WBC, white blood cell.

Comparisons of proportions for those with clinically significant changes in laboratory measures at Week 8 compared to baseline were not possible because, although a few participants (n = 4) had transient elevations in lab measures, no elevations persisted at Week 8. Specifically, one participant in each group experienced transient elevations of AST and ALT at Week 2 follow‐up (XN: AST = 39 U/L and ALT = 81 U/L; Placebo: AST = 69 U/L and ALT = 40 U/L). Values for both participants returned to normal upon re‐testing. One participant in the placebo group also had a transient elevation of AST (37 U/L) and ALT (54 U/L) at Week 6 follow‐up, which returned to normal upon re‐testing. One participant in XN had a transient elevation in serum potassium at Week 4 follow‐up (6.1 mmol/L), which returned to normal upon re‐testing.

3.4. Tolerability

Adherence in both the XN and placebo groups averaged 95%. There were no apparent group‐specific trends in graded, self‐reported symptoms. Both findings suggest favorable tolerability.

3.5. Self‐Reported Disease Activity (CDAI) and Health‐Related Quality of Life (PROMIS‐29)

CDAI trended downward in both groups over the course of the trial. In fact, at Week 8, mean CDAI scores were below 150, which is considered the threshold for clinical remission. There were no statistically or clinically significant between‐group differences in the changes in CDAI between baseline and Week 8. As rates of the induction of remission due to placebo in meta‐analyses of CD clinical trials have been reported to be between 10%–18%, this observation likely represents placebo effect, potentially in both groups. [31, 32]. Figure 2 details the CDAI results.

FIGURE 2.

Change in Crohn's disease activity index.

Changes in PROMIS health‐related quality of life domain scores varied between 5 and 20. Given that PROMIS is population‐normalized with 50 being the average score in all domains, some of these changes are clinically meaningful. Of note, domains most likely affected by Crohn's status, that is, physical interference, fatigue, and social functioning, changed the most dramatically during the trial. However, between group differences in PROMIS domains universally centered around “0” (no difference), and there were no statistically significant differences between XN and placebo. Figure 3 details the PROMIS‐29 results.

FIGURE 3.

Change in PROMIS health‐related quality of life.

4. Discussion

This trial demonstrates that 24 mg/day XN for 8 weeks is safe and well tolerated in adults with active CD, extending prior findings from healthy volunteers [14]. Given the participants in this trial had mild, but not remitted CD (i.e., average baseline CDAI ∼215) and a relatively high self‐reported quality of life (i.e., all baseline mean PROMIS domain scores were ∼50 or greater), these results may not generalize to those with advanced CD. The favorable safety and tolerability results are particularly relevant given the high incidence of AEs associated with standard CD therapies, which include serious infections, hepatotoxicity, and increased malignancy risk [1, 2]. XN's safety profile, and potentially innovative mechanisms of action in chronic inflammatory bowel disease [10, 13, 33] support its potential as a safe candidate therapeutic in CD. However, the dose studied here is very low, and as such, the findings herein help justify future dosing studies to further investigate the engagement targets of XN and to further evaluate its safety and tolerability at higher doses.

Mechanistically, XN may exert multiple beneficial actions relevant to CD pathogenesis and progression. In pre‐animal models, XN's activation of Nrf2 enhances antioxidant defenses, mitigating oxidative injury to mucosal cells [34]. Further in vitro and non‐human in vivo data support XN activity via FXR modulation, in which XN may reduce bile acid–induced epithelial damage and restore barrier integrity [35, 36]. Further, in vivo evidence supports inhibition of NF‐κB by XN, and subsequent reductions in the transcription of pro‐inflammatory cytokines such as TNF‐α, IL‐1β, and IL‐6 [37]. Our prior Phase 1 clinical trial in healthy adults demonstrated XN reduced microbiota‐derived secondary bile acids and altered gut microbial taxa in an enterotype‐dependent manner, findings that may translate into less bile acid–driven inflammation in CD [10]. These findings combined support multiple potential systemic molecular targets by XN, as well as attenuation of several gut‐specific dysfunctions established in IBD. However, the translation of these findings into humans with IBD is far from complete. In addition to the clinical safety results reported here, additional results from this Phase 2 trial demonstrated that participants with higher baseline inflammation (as defined as TNF‐α, IL‐1β, and IL‐6each > 30 pg/mL) had greater and statistically significant reductions in their CDAI scores compared to those with low inflammation at baseline (Jamieson et al., under review); this finding suggests the possibility of a XN‐response phenotype, though deserves replication. In addition to dose response trials, ideally, endoscopy and biopsy would be performed to evaluate changes in mucosal injury and inflammation, and may be critical before longer‐term Phase 3 efficacy trials will be justifiable.

The observed reductions in both BMI and GGT in the XN group relative to placebo are of potential clinical interest. Although the BMI reduction within the XN group was not particularly noteworthy, the difference between XN and placebo (−0.67 kg/m², p = 0.04) may be clinically meaningful, though the statistical significance of this finding would not withstand correction for multiple comparisons. On one hand, the BMI increase noted in the placebo group could be interpreted as a positive change, corresponding to weight gain in parallel to an improvement in CDAI scores; however, participants were not underweight at baseline. In fact, per Table 1, participants in the XN and placebo groups, on average, would be classified as overweight or obese, respectively. Thus, weight loss, or weight maintenance as in the case of participants in the XN group in this trial, may be advantageous from a cardiometabolic risk perspective.

The findings from this trial related to GGT reductions have more promise and potential clinical significance. BMI and GGT are closely correlated, though GGT adds independent predictive value for adverse cardiometabolic risk [26]. Elevated GGT is associated with oxidative stress, endothelial dysfunction, and increased cardiovascular risk—including among individuals with CD, who have higher rates of CVD than the general population [24, 26]. Related, GGT predicts metabolic dysfunction‐associated steatotic liver disease (MASLD) [27]. XN has demonstrated pre‐clinical evidence of improving glucose and lipid metabolism in obese animals [19] and reducing the hepatic consequences of metabolic dysfunction [11], thus the GGT findings may be causally related to established mechanisms of XN, including its effects on bile metabolism. Notably, even minor increases in GGT within the clinically normal range are associated with increased atherosclerotic risk biomarkers [26] and increased odds of insulin resistance and metabolic disease [38]. Thus, XN may be a novel therapeutic candidate if, in addition to its effects in the gut, it reduces cardiometabolic risk in those with inflammatory bowel disease.

Collectively, these pleiotropic mechanisms support XN's potential to target key drivers of CD pathology: dysbiosis, mucosal inflammation, oxidative stress, and barrier dysfunction [5, 6, 7, 29]. While this phase 2 study was not powered to assess efficacy, the favorable safety and biomarker trends reported here justify further research into XN's therapeutic potential in CD and its associated CVD risk. Our suggestions would include larger dose‐response trials; replication of the engagement targets we've identified in our phase 1 and phase 2 trials to date; further exploration of dose‐response phenotypes, perhaps with selective inclusion of participants with increased baseline inflammation; and ideally, though costly, histologic measures of mucosal inflammation.

Strengths of this study include its masked design, serial evaluations, high adherence, serial measurement of XN stability throughout the trial, and biomarker assessment, including XN metabolites to confirm adherence. Limitations include the small sample size, relatively healthy sample despite having CD (based on near median PROMIS t‐scores at baseline), and modest study duration, which precluded detection of rare AEs, generalizations of safety, and assessment of long‐term effects, respectively. Although we were not able to recruit our targeted number of participants, that is, n = 24, the very low frequency of moderate or greater severity AEs and the absence of any clinically meaningful or persisting changes in laboratory measures reduce concerns of inadequate statistical power, although again, we were not powered to detect rare events. Another key limitation includes only evaluating one, relatively low dose of XN.

5. Concluding Remarks

In conclusion, XN demonstrated excellent safety and tolerability in adults with active CD, with clinical laboratory biomarkers suggesting potential metabolic and hepatic benefits. These findings warrant larger, longer‐term studies to assess XN dose optimization, additional safety assessments, and interim measures of clinical efficacy in IBD.

Conflicts of Interest

R.B. and co‐investigators previously received grant support and study products from Metagenics Inc., but have no financial ties to the outcome of this trial.

Data Availability Statement

Data supporting the findings are available from the corresponding author upon reasonable request.