Advances in Noninvasive Measurement of Liver Function and Physiology: The HepQuant DuO Test

Abstract

Current noninvasive liver tests are surrogates for fibrosis and lack ability to directly measure liver function. HepQuant’s tests measure liver function and physiology through hepatic uptake of stable cholate isotopes. HepQuant SHUNT (V1.0) involves oral and intravenous dosing and six blood samples over 90 minutes. We developed simplified test versions: SHUNT V2.0 (oral and intravenous dosing, two blood samples over 60 minutes) and DuO (oral dosing only, two blood samples over 60 minutes). The aim of this study was to evaluate equivalency of the simplified tests to the original SHUNT test. Data from three studies comprising 930 SHUNT tests were retrospectively analysed by each method. Equivalence was evaluated in terms of proportion of tests in which the difference between methods was less than any clinically meaningful difference, and additionally by two one-sided t-test and bioequivalence methods. DuO and SHUNT V2.0 were equivalent to the original SHUNT test for Disease Severity Index, with >99% and >96% of tests falling within equivalence bounds. DuO and SHUNT V2.0 met equivalency criteria by two one-sided t-tests and bioequivalence. DuO and SHUNT V2.0 are easier to administer, less invasive than the original SHUNT test, and have potential to be more accepted by patients and providers.

Introduction and Background

Chronic liver disease is one of the leading causes of death, accounting for over 51,000 deaths (1.5%) in 2020.¹ Chronic liver disease can result in liver cancer, decompensation, liver failure, and death if left untreated. To assess a patient’s liver health, primary care physicians, hepatologists, and gastroenterologists currently use methods including nonspecific blood tests, biomarker panels, elastography surrogates, liver biopsy, and invasive radiologic procedures—none of which directly measure liver function and physiology.

Elastography, which measures liver stiffness, is commonly used to estimate the extent of fibrosis. However, its reliability is influenced by the operator’s skills, and the relationship between liver stiffness and degree of fibrosis is complicated by factors such as hepatic steatosis, obesity, hepatic congestion, hepatic inflammation, and fasting versus fed status. Additionally, most blood-based biomarkers elucidate hepatic fibrosis rather than liver function. For example, individual components of the ELF^™ (Enhanced Liver Fibrosis) test,² the first and only blood test granted FDA marketing authorization for prognosis of advanced fibrosis due to MASH, may be affected by non-hepatic factors, and none of them directly measure hepatic function.

Currently, therapeutic drug trials often use liver histology and hepatic venous pressure gradient (HVPG) measurements as endpoints. However, these methods are costly, invasive, and impractical for baseline testing or monitoring. Additionally, the endpoints provided by these methods often lack reproducibility and are prone to sampling errors, making them ineffective in detecting responses.³ In clinical practice, liver biopsy and HVPG are not well-tolerated by patients and are prone to the same practical challenges. There is a pressing need for accurate and minimally invasive assessments of liver function to assist healthcare providers, patients, and payers in managing the complete spectrum of liver diseases.

HepQuant is a unique testing platform that is suitable for addressing the spectrum of liver disease etiologies and severities through the evaluation of both hepatocyte function and portal circulation.^4–18 HepQuant tests measure a natural process that is specific to the liver—the uptake of cholate—and is, thus, agnostic to liver disease etiology. The HepQuant test measures serial serum concentrations of orally- and intravenously-administered labeled cholates using highly sensitive liquid chromatography and mass spectrometry. The resulting clearance curves are used to generate liver health/disease indices that are associated with liver function, portal circulation, and risk for clinical outcomes.

Despite the strong promise of the HepQuant SHUNT test, its implementation involves administering both oral and intravenous (IV) cholate doses and collecting six peripheral venous blood samples over a 90-minute period. This process introduces potential sources of error, such as challenges with IV access, the risk of extravasation, timing errors during sample collection, and variability in the administration skills of the test.

In our previous work, we created a compartmental model for the dual cholate clearance assay, enhancing its reproducibility and interpretation of hepatic extraction and portal-systemic shunt.¹⁹ Using the compartmental modelling approach, we developed simplified versions of the test, SHUNT V2.0 and DuO, which require fewer blood draws, reduce the testing time, simplify the test administration, and enhance the reproducibility of test parameters (see Supplemental Materials for description of analysis methods). In this same work we demonstrated that SHUNT V1.1, which eliminates the need for the 5-minute sample, was more reproducible than SHUNT V1.0 and, thus, was designated as the reference method for HepQuant tests. This work investigates whether HepQuant’s DuO and SHUNT V2.0 tests are equivalent to the reference (SHUNT V1.1) method.

Methods

This study is a retrospective analysis of data collected from three different studies of the HepQuant SHUNT test: the HepQuant Reproducibility Study (REPRO),²⁰ the Primary Sclerosing Cholangitis (PSC) Study,¹³ and the Hepatitis C Antiviral Long-Term Treatment Against Cirrhosis Trial (HALT-C) Quantitative Liver Function Test (QLFT) ancillary study.^5–7,21

Subjects

The REPRO study included 48 subjects from three groups (16 controls, 16 metabolic dysfunction-associated steatohepatitis [MASH], and 16 HCV) who underwent three separate SHUNT tests on three different days within a 30-day period, representing a total of 144 SHUNT tests performed in the study. The healthy controls had no prior history of liver disease and normal standard blood tests. MASH was diagnosed based on risk factors, absence of evidence for other liver diseases, and fibrosis stage by liver biopsy or transient elastography. HCV was diagnosed based on confirmed HCV history by nucleic acid testing and METAVIR fibrosis stage by liver biopsy.

The PSC study included 47 patients with PSC who were recruited from hepatology clinics representing the clinical spectrum of PSC. All 47 subjects had a baseline SHUNT test, 46 had a second baseline test to assess reproducibility, and 40 had a follow up test after one year, representing a total of 133 SHUNT tests performed in the study.

Details of the HALT-C QLFT study have been previously published.^5–7,21 SHUNT tests were performed in 277 subjects at baseline, 212 at Year 2, and 164 at Year 4. A total of 653 SHUNT tests were performed in the HALT-C study.

Informed consent was obtained from all subjects. The HALT-C (clinicaltrials.gov, NCT00006164), REPRO (clinicaltrials.gov, NCT01579162), and PSC studies were approved by institutional review boards and conducted according to The Code of Ethics of the World Medical Association (Declaration of Helsinki). The study was conducted in accordance with the Basic & Clinical Pharmacology & Toxicology policy for experimental and clinical studies.²²

HepQuant Test Versions

The HepQuant Test versions are summarized in Table 1 and described below.

Table 1.

Summary of HepQuant test versions and test parameters.

Test Versions	Cholate Isotope Doses / Blood Samples / Testing Duration
SHUNT V1.0	1 IV + 1 oral / 6 samples / 90 min.
SHUNT V1.1	1 IV + 1 oral / 4 samples / 90 min.
SHUNT V2.0	1 IV + 1 oral / 2 samples / 60 min.
DuO	1 oral / 2 samples / 60 min.
STAT	1 oral / 1 sample / 60 min.
Test Parameters	Description
Portal HFR	Clearance of orally administered cholate isotope corrected for body weight
Systemic HFR	Clearance of IV-administered cholate isotope corrected for body weight
DSI	Global liver function score comprising portal and systemic HFRs indexed to healthy controls
SHUNT%	Portal-systemic shunting fraction defined by ratio of systemic to portal HFRs
Hepatic Reserve	Overall hepatic health score comprising portal and systemic HFRs indexed to healthy controls with lean body mass

Abbreviations: DSI, Disease Severity Index; HFR, hepatic filtration rate; IV, intravenous.

• SHUNT V1.0 involves the simultaneous administration of 13C-CA by IV and d4-CA orally and six timed peripheral venous blood samples over 90 minutes. Cholate concentrations in serum are measured by LC/MS, and a noncompartmental analysis⁴ fits IV and oral clearance curves to measurement data calculate the areas under the IV and oral curves (AUC_IV, AUC_Oral).

• SHUNT V1.1 eliminates the 5-minute sample from the calculations. Instead, the 13C-CA is estimated from a total blood volume calculation²³, and the 5-minute d4-CA is approximated by 15% of the 20-minute concentration. This modification was shown to significantly reduce the variability in measuring test parameters that depend on the systemic clearance. For this reason, SHUNT V1.1 is designated as the reference method of the HepQuant test.

• SHUNT V2.0 further simplifies by shortening the testing window and requiring only two blood draws at 20 and 60 minutes. SHUNT V2.0 uses a compartmental model¹⁹ to measure the portal cholate clearance and noncompartmental exponential fits to measure systemic cholate clearance.

• DuO is an oral-only test involving only one oral dose at 0 minutes and two blood samples collected at 20 and 60 minutes. The IV clearance is derived rather than measured, and the derived IV concentrations are then used in the same noncompartmental analysis as SHUNT V2.0.

• STAT is the simplest test method which involves one oral dose and one blood sample at 60 minutes. The value of the STAT test may be reported as either the STAT score (the d4-CA concentration adjusted to 75 kg body weight), the portal HFR estimated by STAT, or DSI estimated by STAT.

HepQuant Test Parameters

HepQuant’s tests return a series of test parameters which have been previously associated with liver function and disease/health status. This study assessed the HepQuant test parameters, summarized in Table 1 and described below, for their equivalency to the reference method:

• Portal hepatic filtration rate (HFR_P) is the portal clearance adjusted for body weight (units mL min⁻¹ kg⁻¹) calculated by Equation 1:

$${HFR}_P=\frac{D_{PO}}{{AUC}_{Oral}\bullet BW}$$ (1)

where D_PO is the oral cholate dose.

• Systemic hepatic filtration rate (HFR_S) is the measured systemic clearance adjusted for body weight (units mL min⁻¹ kg⁻¹) calculated by Equation 2:

$${HFR}_S=\frac{D_{IV}}{{AUC}_{IV}\bullet BW}$$ (2)

where D_IV is the intravenous cholate dose.

• Disease Severity Index (DSI) is a score between 0 and 50 which indicates overall liver function comprising both portal and systemic HFR. DSI is associated with fibrosis stage and clinical stages of cirrhosis²⁴ and is calculated by Equation 3:

$$DSI=A\bullet \sqrt{{\left(\ln \left[\frac{{HFR}_{P,max}}{{HFR}_P}\right]\right)}^2+{\left(\ln \left[\frac{{HFR}_{S,max}}{{HFR}_S}\right]\right)}^2}$$ (3)

where A is a scaling factor and HFR_P,max and HFR_S,max are the upper limits of clearance for healthy controls.

• SHUNT%, or shunt fraction, is the estimated absolute bioavailability of the oral d4-CA dose in systemic circulation estimated by Equation 4. SHUNT% is a direct measurement of the first pass hepatic extraction of cholate which is influenced by portal blood flow and portal-systemic shunt.

$$SHUNT\mathrm{\%}=\frac{{AUC}_{Oral}\bullet D_{IV}}{{AUC}_{IV}\bullet D_{PO}}$$ (4)

• Hepatic Reserve (HR) represents an individual’s overall hepatic health relative to lean healthy controls, with values between 100 (healthy) and 0 (severely impaired). HR is calculated by Equation 5:

$$HR=100-A\bullet \sqrt{{\left(\ln \left[\frac{{HFR}_{P,lean}}{{HFR}_P}\right]\right)}^2+{\left(\ln \left[\frac{{HFR}_{S,lean}}{{HFR}_S}\right]\right)}^2}$$ (5)

where A is a scaling factor and HFR_P,lean and HFR_S,lean are mean HFRs minus one standard deviation from a population of lean controls.

Statistical Analysis

The HepQuant tests are potentially affected by variance in the administration of test compounds and time of sampling, day-to-day fluctuations in gastrointestinal or hepatic physiology, and variations in sample preparation and laboratory analysis. The upper and lower equivalence bounds were specified based on the smallest effect size of interest defined by the 95% confidence interval for within-individual variability from the REPRO Study²⁰ (see Supplemental Table S1 for the equivalence bounds derived for all test parameters and reference methods). To evaluate agreement, correlation plots compared analysis methods with the reference method (SHUNT V1.1) for determination of test parameters using Deming regression fit to assess systematic differences.²⁵ Bland-Altman plots compared test parameters calculated by each analysis method and the reference method for bias and 95% confidence interval (CI).²⁶ The primary performance measure was the proportion of tests falling within the equivalence bounds (i.e., the difference between analysis methods was less than any clinically meaningful difference). The analysis was completed for each study individually and for all datasets combined.

The FDA has also provided guidance on preferred methods to demonstrate equivalence of tests—two one-sided t-tests (TOST)^27,28 and bioequivalence procedures.²⁹ In the TOST procedure the upper and lower equivalence bound was specified based on the smallest effect size of interest (Supplemental Table S1). Two composite null hypotheses were tested. If both one-sided tests are statistically rejected, the observed effect falls within the equivalence bounds and is statistically smaller than any effect deemed worthwhile and considered practically equivalent. To evaluate bioequivalence, the mean ratio and 90% confidence intervals of the next generation tests and reference method was calculated for each study separately and across all studies. The mean ratios and 90% CIs were compared to bioequivalence limits of 80 and 120% for parameters in their original scale (i.e., not log-transformed).

Results

Subject Characteristics

Subject characteristics and HepQuant test parameters at baseline visit are listed in Table 2 and described by each study population below.

Table 2.

Subject characteristics at baseline. Data are displayed as n (%) or mean ± SD.

	REPRO Study*			PSC Study	HALT-C
	Controls	MASH	HCV
No. of subjects	16	16	16	47	277
Age (y)	32.9 ± 12.0	49.8 ± 11.6	55.6 ± 6.7	48.8 ± 12.9	49.9 ± 7.3
Gender (Male)	8 (50.0%)	7 (43.8%)	13 (81.3%)	36 (76.6%)	209 (75.5%)
BMI (kg/m2)	23.0 ± 2.2	32.5 ± 5.9	28.2 ± 4.1	26.2 ± 3.9	29.4 ± 4.8
Race/ethnicity
Non-Hispanic White	12 (75.0%)	14 (87.5%)	14 (87.5%)	43 (91.5%)	210 (75.8%)
Hispanic	2 (12.5%)	2 (12.5%)	1 (6.3%)	1 (2.1%)	35 (12.6%)
Black	1 (6.3%)	0 (0%)	1 (6.3%)	3 (6.4%)	28 (10.1%)
Other or Unknown	1 (6.3%)	0 (0%)	0 (0%)	0 (0%)	4 (1.4%)
Fibrosis stage
F0	NA	0 (0%)	3 (18.8%)	NA	0 (0%)
F1	NA	4 (25.0%)	2 (12.5%)	NA	0 (0%)
F2	NA	4 (25.0%)	3 (18.8%)	NA	28 (10.1%)
F3	NA	5 (31.3%)	4 (25.0%)	NA	83 (30.0%)
F4	NA	3 (18.8%)	4 (25.0%)	NA	57 (20.6%)
F5	NA	0 (0%)	0 (0%)	NA	51 (18.4%)
F6	NA	0 (0%)	0 (0%)	NA	58 (20.9%)
Fibrosis score	NA	2.4 ± 1.1	2.3 ± 1.5	NA	2.9 ± 1.04
MELD score	NA	7.0 ± 1.1	7.9 ± 2.0	9.5 ± 4.2	6.9 ± 1.3
CTP score	NA	5.0 ± 0.0	5.4 ± 0.7	6.0 ± 1.5	5.3 ± 0.5

Abbreviations: B, Black race; BMI, body mass index; CTP, Child-Turcotte-Pugh; HALT-C, Hepatitis C Antiviral Long-Term Treatment Against Cirrhosis Trial Quantitative Liver Function Ancillary Study; HCV, chronic hepatitis C; MASH, metabolic dysfunction-associated steatohepatitis; MELD, model for end-stage liver disease; NA, not applicable—healthy controls did not undergo liver biopsy and none had clinical disease and fibrosis stage/score was not reported in the PSC study; PSC, Primary Sclerosing Cholangitis Study; REPRO, HepQuant Reproducibility Study.

^*Fibrosis stage was METAVIR in HCV cases; BRUNT/KLEINER in MASH cases; 2 MASH cases had F3 fibrosis by transient elastography; MELD and CTP scores were calculated for the F3/F4 subgroup only.

REPRO Study.

The REPRO Study involved 48 participants divided into three groups: controls (n=16), MASH (n=16), and HCV (n=16). The control group had a mean age of 33 ± 12 years, an equal distribution of males and females (8:8), and a normal BMI of 23 ± 2 kg/m². The MASH group exhibited a mean age of 50 ± 12 years, male-to-female ratio of 7:9, and elevated BMI of 33 ± 6 kg/m². The HCV group had mean age of 56 ± 7 years, male-to-female ratio of 13:3, and moderate BMI of 28 ± 4 kg/m². The majority of participants (83%) were non-Hispanic white. Alanine aminotransferase (ALT), aspartate aminotransferase (AST), gamma-glutamyl transferase (GGT), and glucose were elevated in MASH and HCV relative to controls. Platelets were significantly lower in MASH and HCV. Both MASH and HCV groups had 8 F0-F2 (early) fibrosis and 8 F3-F4 (advanced) fibrosis.

PSC Study.

Out of the 47 subjects in the PSC study, 92% were non-Hispanic white, 6% were African American, and 2% identified as Hispanic. The majority of subjects were male (36:11 M:F) with mean age of 49 ± 13 years and BMI of 26 ± 4 kg/m². Seventy percent of subjects had inflammatory bowel disease, 53% had ulcerative colitis, 36% had history of varices, ascites, or encephalopathy, 36% had splenomegaly, 28% had low platelet count (<140,000 μL⁻¹), 19% had history of jaundice, and 38% had history of bacterial cholangitis. Seventy-two percent of those with ERCP reports (28 out of 39) had undergone stricture dilation, stenting, or both. Sixty-six percent of subjects were treated with UDCA, and 40% were taking medication for inflammatory bowel disease.

HALT-C QLFT Study.

Out of the 277 chronic HCV patients with baseline SHUNT tests in the HALT-C Trial QLFT ancillary study, about 60% had Ishak fibrosis stages 2–4 (fibrosis) and 40% had stage 5 or 6 (cirrhosis). No subjects had experienced clinical decompensation. The majority of subjects were male (76%) with a mean age of 50 ± 7 years and mean BMI of 29 ± 5 kg/m². Ninety-two percent had HCV genotype 1 and mean HCV-RNA of 6.4 ± 0.6 log₁₀ of IU/mL. The mean platelet count was 169,000 ± 68,000 μL⁻¹.

Equivalency to SHUNT V1.1

The equivalence of SHUNT V2.0, DuO, and STAT relative to SHUNT V1.1 are summarized in Table 3 which shows the percentage of tests that fall within the equivalence bands for each study individually and across all studies. Equivalency in DSI between simplified test versions and SHUNT V1.1 was further assessed by TOST and bioequivalence methods (Figure 1). A secondary analysis using SHUNT V1.0 as the reference method was also performed (see Supplemental Table S2 and Supplemental Figures S1–S4).

Table 3.

Equivalence of next generation test methods to SHUNT V1.1 in terms of the number of tests (% of tests) falling within the equivalence bounds.

Parameter	Test Version	REPRO (n=144)	PSC (n=133)	HALT-C (n=653)	Combined (n=930)
DSI	SHUNT V2.0	144 (100) *	132 (99) *	647 (99) *	923 (99) *
	DuO	144 (100) *	131 (99) *	623 (95.4) *	898 (97) *
	STAT	112 (78)	117 (88)	565 (87)	794 (85)
SHUNT% (%)	SHUNT V2.0	144 (100) *	132 (99) *	639 (98) *	915 (98) *
	DuO	133 (92)	103 (77)	554 (85)	790 (85)
	STAT	NA	NA	NA	NA
Hepatic Reserve (%)	SHUNT V2.0	144 (100) *	131 (99) *	625 (96) *	900 (97) *
	DuO	144 (100) *	129 (97) *	606 (93)	879 (95) *
	STAT	NA	NA	NA	NA
Portal HFR (mL min−1 kg−1)	SHUNT V2.0	144 (100) *	132 (99) *	645 (99) *	921 (99) *
	DuO	144 (100) *	132 (99) *	645 (99) *	921 (99) *
	STAT	137 (95) *	131 (99) *	639 (98) *	907 (98) *
Systemic HFR (mL min−1 kg−1)	SHUNT V2.0	144 (100) *	133 (100) *	653 (100) *	930 (100) *
	DuO	111 (77)	101 (76)	555 (85)	767 (82)
	STAT	NA	NA	NA	NA

^*Above acceptable limit (≥95%)

DSI = Disease Severity Index; HFR = hepatic filtration rate; REPRO = HepQuant Reproducibility Study; PSC = Primary Sclerosing Cholangitis study; HALT-C = Hepatitis C Antiviral Long-Term Treatment Against Cirrhosis Trial.

Figure 1.

Equivalence of next-generation tests to SHUNT V1.1 in terms of Disease Severity Index (DSI) by each study individually and across all studies. (A) Two one-sided t-test (TOST) analysis of SHUNT V2.0 for DSI. (B) Bioequivalence analysis of SHUNT V2.0 for DSI. (C) TOST analysis of DuO for DSI. (D) Bioequivalence analysis of DuO for DSI. Error bars represent 90% confidence intervals. Dotted red line indicates equivalence bounds (TOST) or bioequivalence limits. REPRO = HepQuant Reproducibility Study; PSC = Primary Sclerosing Cholangitis Study; HALT-C = Hepatitis C Antiviral Long-Term Treatment Against Cirrhosis Trial; ALL = all studies combined.

SHUNT V2.0.

SHUNT V2.0 was equivalent to the reference method for all test parameters with the percentage of tests within the equivalence bounds ranging from 97% to 100%. The DSI measured by SHUNT V2.0 had excellent agreement with the reference method in terms of DSI (Figure 2), with low bias (0.21 DSI units) and all but seven tests falling within the equivalence bounds and Deming regression line approaching identity (R² = 0.98, slope = 0.99, intercept = −0.05). The DSI from SHUNT V2.0 was also statistically equivalent to, not different from, and bioequivalent to the DSI measured from SHUNT V1.1, across all studies individually and combined (Figure 1). The SHUNT% measured by SHUNT V2.0 also had excellent agreement with the reference method with mean bias of 1.3% and R² of 0.96 (Supplemental Figure S5).

Figure 2.

Equivalence analysis of SHUNT V2.0 versus SHUNT V1.1 for Disease Severity Index (DSI). Bland-Altman plot (left) shows the mean of methods versus the difference between methods with bias (solid black line), 95% confidence intervals of the mean difference (dotted lines), and equivalence bands (red lines). The correlation plot (right) shows the DSI values for both methods and Deming regression (solid line). Each point represents values derived from a single test across all studies in the analysis. HALT-C = Hepatitis C Antiviral Long-Term Treatment Against Cirrhosis Trial; PSC = Primary Sclerosing Cholangitis Study; REPRO = HepQuant Reproducibility Study; MASH = metabolic dysfunction-associated steatohepatitis; HCV = hepatitis C virus.

DuO.

DuO was equivalent to the reference method for DSI (97%), Portal HFR (99%), and Hepatic Reserve (95%)—but not SHUNT% (85%) and Systemic HFR (82.5%). The DSI from DuO showed excellent agreement with the reference method (Figure 3), with Deming regression line near identity (R² = 0.95, slope = 1.01, intercept = −0.16). The DSI from DuO was found to be statistically equivalent, not different, and bioequivalent to SHUNT V1.1 albeit with slightly larger 90% confidence intervals than those for SHUNT V2.0 (Figure 1). SHUNT% estimated by DuO had modest agreement to the reference method (Supplemental Figure S6) (R² = 0.86, slope = 0.87, intercept = 0.004). DuO demonstrated excellent agreement for Hepatic Reserve (Supplemental Figure S7) (R² = 0.97, slope = 1.01, intercept = 0.01). Additionally, Hepatic Reserve from DuO was found to be statistically equivalent, not different, and bioequivalent to SHUNT V1.1 by the TOST and bioequivalence methods (Supplemental Figure S8).

Figure 3.

Equivalence analysis of DuO versus SHUNT V1.1 for Disease Severity Index (DSI). Bland-Altman plot (left) shows the mean of methods versus the difference between methods with bias (solid black line), 95% confidence intervals of the mean difference (dotted lines), and equivalence bands (red lines). The correlation plot (right) shows the DSI values for both methods and Deming regression (solid line). Each point represents values derived from a single test across all studies in the analysis. HALT-C = Hepatitis C Antiviral Long-Term Treatment Against Cirrhosis Trial; PSC = Primary Sclerosing Cholangitis Study; REPRO = HepQuant Reproducibility Study; MASH = metabolic dysfunction-associated steatohepatitis; HCV = hepatitis C virus.

STAT.

For the STAT test, Portal HFR (98%) was equivalent to the reference method, but not DSI (85%). The DSI estimated by the STAT test (Supplemental Figure S9) had a relatively low bias (−0.93 DSI units), but its overall agreement with the reference method was modest in terms of correlation (R² = 0.89, slope = 0.92, intercept = 2.52).

Discussion

This study used several methods to confirm the equivalency of DSI from the HepQuant DuO and SHUNT V2.0 tests to DSI from the more complicated SHUNT test (V1.0/1.1). Table 4 summarizes the key findings from this study as well as the findings from our reproducibility analysis (submitted for publication, 2023). DSI from SHUNT V1.0 has been associated with a number of relevant laboratory tests, clinical models, clinical complications, and clinical outcomes. By demonstrating equivalency to DSI from SHUNT (V1.0/1.1), as well as high within-individual reproducibility, DSI from DuO or SHUNT V2.0 is likely to also demonstrate the same associations with clinical endpoints.

Table 4.

Summary of equivalence testing of simplified methods (SHUNT V2.0 and DuO) relative to SHUNT V1.1 and reproducibility (submitted for publication) by whether the criteria were met (Y=yes, N=no).

Parameter	Test Version	Equivalency to SHUNT V1.1			Reproducibility
		≥ 95%	TOST	BE	ICC>0.7
DSI	SHUNT V2.0	Y	Y	Y	Y
	DuO	Y	Y	Y	Y
SHUNT%	SHUNT V2.0	Y	Y†	Y	Y
	DuO	N	Y†	N	Y
Hepatic Reserve	SHUNT V2.0	Y	Y	Y	Y
	DuO	Y	Y	Y	Y
Portal HFR	SHUNT V2.0	Y	Y†	N	Y
	DuO	Y	Y†	N	Y
Systemic HFR	SHUNT V2.0	Y	Y	Y	Y
	DuO	N	Y†	N	Y

^†Statistically equivalent and statistically different (i.e., the difference is not meaningful in this application).

For equivalency, the criteria were ≥ 95% of tests within equivalence bounds, two one-sided t-test (TOST), and bioequivalence (BE). For reproducibility, the criterion was intraclass correlation coefficient (ICC) > 0.7. DSI = Disease Severity Index; HFR = hepatic filtration rate.

When treating organ-specific diseases, it is critical to establish the functional capacity of that organ. Functional tests are available for major organs: the kidneys (creatinine clearance), lungs (spirometry, gas exchange, plethysmography), and heart (ejection fraction). With no commercially available organ function test for the liver, care providers are limited to alternative methods to interrogate the liver (liver enzymes, general lab tests, liver biopsy, biomarkers, liver stiffness, HVPG). While these tests provide valuable information about different aspects of liver disease, none of these tests provide a true readout of liver function.

HepQuant fulfills a significant unmet need as a functional test of the liver. The HepQuant SHUNT test is a comprehensive diagnostic tool that considers both hepatocyte function and portal circulation, addressing the pathogenesis across a spectrum of liver disease severities and etiologies, including cirrhosis,¹⁷ hepatitis C virus (HCV),^6,7,16 primary sclerosing cholangitis (PSC),¹³ hepatocellular carcinoma,¹² portal hypertension,^10,11 and MASH.¹⁴ A recent pivotal study demonstrated a correlation between DSI and the probability of developing large esophageal varices.^30–32

In SHUNT V1.0, the initial concentration at time 0 minutes (C₀) is estimated from ln/linear regression of [13C-CA] versus time using the 5- and 20-minute time points. Minor variations in timing the 5-minute sample or failure to clear injected 13C-CA from the line in the one arm method introduce errors that add variance to parameter determination, particularly volume of distribution (Vd).

Previously, we showed that SHUNT V1.1, which is identical to SHUNT V1.0 except for elimination of the 5-minute time point, was more reproducible than SHUNT V1.0 across nearly every test parameter, with significant improvement in the reliability of SHUNT%. We evaluated the basis for this improved reproducibility in a subset of 179 subjects who underwent two-arm catheter administration/collection from the SHUNT-V study.³² The mean Vd measured by SHUNT V1.1 (0.058 ± 0.007 L/kg) was identical to the mean Vd measured by SHUNT V1.0 (0.058 ± 0.018 L/kg). However, the variance in Vd from SHUNT V1.1 was less than half of the variance in Vd from SHUNT V1.0. These results imply that elimination of the 5-minute time point and calculation of C₀ based on BMI²³ is not only justified but, by reducing variance, improves test performance.

In this study, we demonstrated that SHUNT V1.1 was equivalent to SHUNT V1.0 with >98% of values within the equivalence bounds for key test parameters. Based on equivalency to SHUNT V1.0 and improved reproducibility, we recognize SHUNT V1.1 as the reference method for comparing the next-generation tests (SHUNT V2.0 and DuO).

This study investigated whether SHUNT V2.0, which reduces the testing duration from 90 to 60 minutes and requires only two blood samples, could provide practically equivalent measurements of test parameters relative to SHUNT V1.1. Previously in the development of the analysis for SHUNT V1.0, a 2-point version of the noncompartmental analysis was not reproducible or accurate.⁴ In contrast, SHUNT V2.0, which uses a compartmental analysis of portal clearance to model aspects of liver function and physiology, was practically equivalent to SHUNT V1.1 with near perfect agreement in terms of test parameters falling within the equivalence bounds. Compared to SHUNT V1.1, SHUNT V2.0 offers a simpler and less invasive administration, making it potentially more widely embraced by healthcare providers who administer the test and by patients undergoing the test. The SHUNT V2.0 test is highly applicable in clinical situations that require accurate and precise assessments of systemic clearance, such as the monitoring of portal-systemic shunting or the evaluation of novel treatments that may affect systemic clearance.

We also explored whether a further simplified test using only oral cholate dosing (DuO) could determine test parameters with acceptable equivalency to SHUNT V1.1. DuO will alleviate potential errors associated with intravenous (IV) injection, such as challenges in locating a suitable vein for IV access, catheters becoming dislodged from the vein, and the risk of extravasation. Additionally, DuO has the potential to decrease variability linked to the skill required for test administration and the collection of multiple timed blood samples. We found that DuO was equivalent to the reference method for several test parameters which emphasize portal clearance (DSI and Portal HFR), suggesting that DuO could prove valuable in clinical situations that prioritize precise measurements of portal clearance and comprehensive liver function assessment. With its exclusively oral dosing method, DuO has the potential to expand market accessibility to facilities that lack phlebotomy capabilities or enable convenient test administration at home (e.g., through transcutaneous sample collection).

Variable gastric emptying and delayed absorption of the oral dose beyond 60 minutes could invalidate the oral clearance curves. This would manifest as very low to undetectable levels of cholate isotopes which would be flagged by the lab during quality control and designated as an invalid result. Repeat testing would be required. Current experience suggests that this might occur in approximately 1% of tests.

We analyzed DuO and SHUNT simplified methods retrospectively from data that was collected using the SHUNT V1.0 method of test administrations. This study design is ideal since the DuO test and other simplified SHUNT versions are embedded within the SHUNT test, thus avoiding day-to-day variances when tests are performed on separate days or at different times. In addition, the LCMS assay has been extensively validated for ability to independently and accurately quantify individual cholate isotopes when both are co-administered. In fact, it is likely that the more complicated tests (SHUNT V1.0 and SHUNT V1.1) are more susceptible to variability across sites, test administrators, and timing of blood samples and, thus, equivalency of the simplified tests (SHUNT V2.0 and DuO) may be underestimated in the current study.

Clinical validation of DuO and SHUNT V2.0 is ongoing, for example, in the prediction of clinical outcomes in PSC, diagnostic performance of DSI for ruling out large esophageal varices, predicting drug PK in hepatic impairment studies, and monitoring treatment effects in drug trials. Further prospective validation of the simplified test administration is planned.

In conclusion, DuO and other simplified versions of the prototypical SHUNT V1.0 test represent advances in the noninvasive measurement of human liver function and physiology. DuO and SHUNT V2.0 are simple to administer liver function tests that will aid providers in defining severity of liver disease, tracking disease progression, monitoring treatment effects, and communicating this objective information to their patients.

Data Availability Statement

Research data are not shared.