Oxidative breakdown of octanoic acid is maintained in patients with cirrhosis despite advanced disease

Abstract

Abstract  As an octanoic acid ¹³CO₂ breath test is frequently used to test gastric emptying of solid food, the purpose of the present study was to study whether oxidative breakdown of octanoic acid is affected by severe liver disease. The design of our study was twofold. First, cirrhotic patients (n = 82) of varying severity were compared with healthy controls (n = 17). Values of half-time, time point of maximal expiration and cumulative recovery of octanoic acid breath tests (OBT) were not significantly different between them. Secondly, cirrhotic patients (n = 10) were studied before placement of transjugular intrahepatic portosystemic shunt, 4–7 days later and 1–2 months later. Values of half-time, time point of maximal expiration and cumulative recovery of consecutive OBTs did not change significantly. The OBT may therefore be a suitable test in the future to detect delayed gastric emptying of solids in cirrhotic patients with reduced liver function and portal hypertension.

Introduction

Octanoic acid is a medium-chain fatty acid consisting of eight carbon atoms. The ¹³C-octanoic acid breath test (OBT) has been validated and used to measure gastric emptying rate of solids.^1,2 This ¹³CO₂-breath test is a non-invasive and safe gastric motility test and can be easily repeated over short periods of time.^1,2 The use of OBT for gastric motility disorders depends on gastric emptying rate which is the rate-limiting step of this test. It is largely unknown whether liver diseases affect metabolic handling of octanoic acid. The impact of liver impairment on the OBT is relevant in view of possible gastric motility problems observed in patients with cirrhosis and/or portal hypertension.^3,4 OBT ‘baseline’ values for cirrhotic patients are needed.

Octanoic acid has the advantage that it is rapidly absorbed, oxidized and eliminated as CO₂ in the breath. The intestinal mucosa absorbs octanoic acid without previous incorporation in micelles.⁵ Non-esterified and albumin bound, octanoic acid is transported via the portal vein directly to the liver.^5,6 Hepatocytes preferentially oxidize it to acetyl-CoA and CO₂ in mitochondria⁷ and, subsequently, CO₂ is eliminated in the breath.^8,9 Orally administered long-chain fatty acids exhibit a far more complex route and metabolic handling and are therefore less traceable in the breath.^8–10 Not surprisingly, hepatic uptake and oxidation of octanoic acid measured by positron emission tomography, was proposed as a new liver function test and imaging tool.¹¹

The aim of the present study therefore was to investigate whether oxidative breakdown of octanoic acid may be affected in patients with cirrhosis and in portosystemic shunting.

Material and methods

The design of our study was twofold. First, we compared the values of the OBT obtained in a large group (n = 82) of patients with varying severity of cirrhosis, with those of healthy controls (n = 17). The degree of liver impairment was quantified by the Child–Pugh score and assessed by the aminopyrine breath test (ABT).¹² Secondly, a prospective follow-up study was carried out in 10 cirrhotic patients undergoing a transjugular intrahepatic portosystemic shunt (TIPS) procedure. The effects of enhanced shunting as a result of TIPS on the OBT were studied.

¹³C-octanoic acid breath test

After an overnight fast, 91 mg of (1,¹³C-)octanoic acid, obtained from Eurisotop (St Aubin, France), was given orally in 100 ml of water with polyethyleneglycol added to solubilize the lipophilic marker. Octanoic acid was administered orally in a small liquid volume in order to avoid possible gastric motility disturbance with solids. Breath samples were obtained before intake and subsequently every 15 min during 4 h.¹³ The ¹³CO₂ content in the breath was measured with isotope ratio mass spectrometry as published previously.^2,13 The percentage dose recovery over time was fitted with a previously published mathematical model.¹³ OBT_c is the cumulative exhaled amount of ¹³CO₂-tracer calculated for the theoretical time point +∞ and is expressed as % recovery of the dose administered.¹³ OBT half-time (t_1/2) is the time point when 50% cumulative ¹³CO₂ expiration is achieved and is expressed in minutes.¹³ OBT t_max is the time point of maximum exhaled ¹³CO₂-tracer and is expressed in minutes.¹³

¹⁴C-aminopyrine breath test

The ¹⁴C-ABT is a microsomal liver function test.¹² All patients with liver disease underwent an ABT simultaneously with the ¹³C-OBT. ¹⁴C-dimethyl-aminopyrine was given orally (5 μCi), immediately following ¹³C-octanoic acid intake. Breath samples for ¹⁴CO₂ were collected before intake and every 30 min afterwards for 2 h. ¹⁴CO₂ content was measured in a β-scintillation counter (Tri-Carb 2100 TR; Packard, Downers Grove, IL, USA). ABT 2 h is the measured cumulative recovery after 2 h and is expressed as a percentage of the administered dose.

Healthy volunteers

Seventeen healthy volunteers (eight male, nine female), 23 ± 5 years (mean ± SD) underwent the ¹³C-OBT in the fasted state. They did not take any medication in the week before or on the day of the breath test. Control values for the ¹⁴C-ABT had been obtained before in our laboratory.

Patient groups

None of the patients studied had severe diabetes mellitus and/or complaints of gastric emptying problems.

Cross-sectional study

Eighty-two patients with cirrhosis were studied. The aetiology of the disease was alcohol (n = 36), alcohol and hepatitis C (n = 6), hepatitis C (n = 7), hepatitis B (n = 8), biliary cirrhosis (n = 12), α₁-antitrypsin deficiency (n = 2), haemochromatosis (n = 2), Budd–Chiari (n = 1), autoimmune hepatitis (n = 1) or cryptogenic cirrhosis (n = 7). All diagnoses were documented by liver histology and appropriate serum tests and imaging. The patients were classified in Child–Pugh classes:¹⁴ A (n = 7), B (n = 30) or C (n = 45). Patient characteristics according to the Child–Pugh classes are given in Table 1.

Table 1.  Characteristics of cirrhotic patients

	Child–Pugh class
(n = 82)	A (n = 7)	B (n = 30)	C (n = 45)
Characteristics of cirrhotic patients in the cross-sectional study, divided according to Child–Pugh classification. Data are given as median and (range). *P < 0.01 as compared with values in Child–Pugh classes A and B. ALT: alanine aminotransferase; AST: aspartate aminotransferase.
Age (mean ± SD years)	53 ± 13	54 ± 10	53 ± 11
Male/female	6/1	22/8	32/13
Haemoglobin (g dl^−1)	11.9 (8.9–16.1)	12.0 (7.4–14.7)	11.8 (4.6–16.1)
Prothrombin time (%)	86 (53–100)	71 (44–100)	47 (19–64)^*
Plasma albumin (g l^−1)	36 (30–46)	34 (17–43)	29 (22–41)^*
ALT (U l^−1)	45 (20–180)	27 (9–187)	42 (3–499)
AST (U l^−1)	61 (15–177)	43 (13–253)	66 (8–609)^*
Alkaline phosphatase (U l^−1)	321 (192–1472)	336 (158–1296)	311 (157–1684)
γ -Glutamyltransferase (U l ^−1 )	89 (40–697)	89 (20–698)	65 (15–398)
Bilirubin (mg dl^−1)	1.1 (0.5–1.9)	1.4 (0.3–9.6)	3.1 (0.9–30.1)^*
Creatinine (mg dl^−1)	0.9 (0.7–1.2)	1.0 (0.6–8.3)	1.0 (0.6–3.5)
Platelet count (10^9 l^−1)	108 (37–412)	124 (13–369)	109 (20–293)

Prospective study in TIPS patients

In ten patients undergoing an elective TIPS procedure, OBT and ABT were performed the day before the TIPS placement, 4–7 days later and 1–2 months later. Eight of 10 patients were male and their age was 52 y ± 13 years (mean ± SD). The indication for elective TIPS placement was treatment of refractory ascites (n = 9) or prophylaxis of rebleeding (n = 1). The aetiology of cirrhosis was alcohol (n = 6), alcohol and hepatitis C (n = 1), hepatitis C (n = 2) and Budd–Chiari (n = 1), all of which were histologically confirmed. Five patients belonged to Child–Pugh class B and five to class C. Oesophageal varices were present in eight of 10 patients. Biochemical data are given in Table 2. The TIPS was created and dilated to achieve a reduction of the porta-caval pressure gradient of at least 50%.¹⁵ In the first week after TIPS placement, an angiographic control was carried out in all patients to document patency of the TIPS. During this control, additional dilatation and/or stenting was performed in three of 10 patients and in one patient recanalization was carried out for a thrombus in the TIPS.¹⁵ Residual porta-caval pressure gradients were then between 2 and 8 mmHg in all. At the time of invasive TIPS control 1–2 months later, no additional angioplasty had to be carried out, because all porta-caval pressure gradients were between 2 and 10 mmHg. OBT and ABT were always carried out before the angiographic procedure of TIPS placement or TIPS control.

Table 2.  Characteristics of TIPS patients

(n = 10)	Before TIPS placement	4–7 days after TIPS placement	1–2 months after TIPS placement
Biochemical parameters and encephalopathy grading in ten cirrhotic patients before and after TIPS placement. Data are given as median and (range). *P < 0.05 as compared with value before TIPS placement. **P < 0.05 as compared with value 1–2 months after TIPS placement. ALT: alanine aminotransferase; AST: aspartate aminotransferase.
Haemoglobin (g dl^−1)	11.5 (7.8–16.1)	10.8 (8.7–13.0)	10.8 (6.9–14.5)
Prothrombin time (%)	66 (42–95)	55 (39–88)	58 (28–95)
Plasma albumin (g l^−1)	31 (23–36)	29 (23–38)	32 (22–42)
ALT (U l^−1)	26 (3–50)	114 (24–348)^*,**	29 (4–84)
AST (U l^−1)	47 (8–109)	98 (41–454)^*	56 (15–143)
Bilirubin (mg dl^−1)	1.2 (0.7–4.8)	1.7 (0.8–6.9)	1.9 (1.0–8.8)^*
Creatinine (mg dl^−1)	1.1 (0.7–4.3)	0.8 (0.6–4.4)	0.9 (0.6–5.3)
Platelet count (10^9 l^−1)	133 (60–262)	127 (63–163)	98 (74–151)
Encephalopathy grading  (scoring 1–3)	1 (1–2)	2 (1–3)	1 (1–3)

All patients and healthy volunteers gave their informed consent before entering the study. The study protocol was approved by the ethical committee of the Faculty of Medicine of Leuven, Belgium.

Biochemical analyses

Routine biochemical analyses on blood/plasma (Tables 1 and 2) were performed with automated standardized procedures (Roche Hitachi 917/747; Roche, Mannheim, Germany). Reference values for haemoglobin are 14.0–18.0 g dl⁻¹ for males and 13.0–16.0 for females, prothrombin time 70–100%, plasma albumin 35–52 g l⁻¹, alanine aminotransferace (ALT) 5–37 U l⁻¹, aspartate aminotransferase (AST) 5–40 U l⁻¹, alkaline phosphatase 90–260 U l⁻¹, γ-glutamyltransferase 11–50 U l⁻¹, bilirubin 0.2–1.0 mg dl⁻¹, creatinine 24–170 mg dl⁻¹ and platelet count 150–450 × 10⁹ l⁻¹.

Statistics

Endpoints of the study were OBT t_1/2, OBT t_max and OBT_c. As most of the data in this study were not normally distributed, data are given as median and range (unless specified otherwise). Two groups were compared with a Wilcoxon rank test. To compare repeated measurements in one group, we used repeated measures analysis of variance on ranks (Friedman's test). Non-parametric correlations were calculated with the Spearman rank order test. Significance levels were always taken at P < 0.05.

Results

Patients with cirrhosis vs healthy controls

The OBT t_1/2 was not significantly different between healthy controls and cirrhotic patients: median (range): 88 (60–135) and 96 (68–230) min, respectively (P = 0.19) (Fig. 1a and Table 3). OBT t_max was not significantly different: 38 (21–86) and 40 (14–88) min, respectively (P = 0.65) (Table 3). OBT_c was not significantly different neither: 45 (29–80) and 55 (23–100)% dose, respectively (P = 0.38) (Table 3). The ¹³CO₂ expiration curves were reconstructed for each group in Fig. 2, using upper limit (upper line) and lower limit (lower line) of the 95% confidence interval of OBT_c. The OBT curve shape showed a small difference in the tail, which was ‘lifted up’ in the liver disease group (Fig. 2). As to aminopyrine, there was a significant difference in ABT 2 h between the Child–Pugh classes (P < 0.01; Table 3). This reflects an impaired microsomal demethylation of aminopyrine in the patients with the most advanced cirrhosis (Fig. 1b).¹²

Figure 1. (a) Comparison of a healthy control group ( n  = 17) with a large group of cirrhotic patients of various aetiology ( n  = 82). Individual data of the half-time in the octanoic acid breath test (OBT t_1/2 ), expressed in minutes, are given. OBT t_1/2 values of cirrhotic patients are presented according to the Child–Pugh score, ¹⁴ ranging from 5 to 15. There was no statistically significant difference between healthy controls and cirrhotic patients ( P  = 0.19). (b) In cirrhotic patients of various aetiology ( n  = 82), individual 2 h cumulative values of the aminopyrine breath test (ABT) are plotted according to the Child–Pugh score, ¹⁴ ranging from 5 to 15. ABT 2 h is expressed in % dose. In our laboratory, normal values averages 18% dose with range minimum of 13%. With increasing Child–Pugh score, ABT 2 h values lowered (see also Table 3 ).

Table 3.  OBT and ABT in cirrhotic patients

	Healthy controls (n = 17)	Cirrhotic patients (n = 82)	Child–Pugh class
			A (n = 7)	B (n = 30)	C (n = 45)
Results of the octanoic acid breath tests (OBT) and aminopyrine breath tests (ABT) in healthy controls and cirrhotic patients in the cross-sectional study. Data are given as median and (range). *P < 0.01 as compared with values in Child–Pugh classes A and B. ND: not done.
OBT t1/2 (min)	88 (60–135)	96 (68–230)	85 (68–117)	98 (69–176)	100 (70–230)
OBT tmax (min)	38 (21–86)	40 (14–88)	38 (23–51)	38 (14–77)	45 (21–88)
OBTc (% dose)	45 (29–80)	55 (23–100)	49 (36–59)	51 (39–78)	57 (17–100)
ABT 2 h (% dose)	ND	2 (0–20)	6 (3–20)	3 (0–15)	1 (0–20)^*

Figure 2. The ¹³CO₂ octanoic acid breath test (OBT) expiration curves were reconstructed for the healthy control group (dotted line) and the cirrhotic patients group (full line), using the group's median OBT t_1/2, the group's median OBT t_max, and upper limit (upper line), respectively, lower limit (lower line) of the 95% confidence interval of cumulative OBT.

In cirrhotic patients, OBT t_1/2 was not correlated with Child–Pugh score, plasma albumin, bilirubin or prothrombin time (Table 4) or other biochemical data (not shown). OBT t_1/2 did correlate negatively with ABT 2 h (Table 4).

Table 4.  Correlations of OBT t_1/2 with parameters of liver function in cirrhotic patients

Octanoic acid breath test	Liver function parameter	Correlation coefficient (n = 82)	P -value
Correlation coefficients of octanoic acid breath test half-time (OBT t1/2) with regard to liver function parameters, together with their P-value.
OBT t1/2 (min)	Child–Pugh score (5–15)	0.092	0.47
OBT t1/2 (min)	Plasma albumin (g l^−1)	0.041	0.71
OBT t1/2 (min)	Plasma bilirubin (mg dl^−1)	0.039	0.73
OBT t1/2 (min)	Prothrombin time (%)	−0.245	0.06
OBT t1/2 (min)	ABT 2 h (% dose)	−0.330	0.005

Effect of the TIPS procedure on OBT and ABT

The parameters OBT t_1/2, OBT t_max and OBT_c did not change significantly after TIPS placement (Table 5), although a tendency was found on a delayed OBT t_max 1–2 months after TIPS placement (P = 0.06; Table 5). The latter value was within the range of t_max in healthy controls, however. We observed a significant delay in aminopyrine demethylation following TIPS procedure as seen by ABT 2 h (P = 0.04; Table 5). Serum levels of bilirubin gradually increased from 1.2 (0.6–4.8) before TIPS to 1.9 (1.0–8.8 mg dl⁻¹ 1–2 months after TIPS procedure (P = 0.02; Table 2). Seven of 10 TIPS patients developed mild encephalopathy or had a worsening of their pre-TIPS situation (Table 2), which necessitated lactulose treatment. In the first week after TIPS placement, serum transaminase activities were significantly higher and returned to baseline values 1 month later (Table 2). Three patients died and one underwent a liver transplantation within the year after TIPS placement.

Table 5.  OBT and ABT in TIPS patients

(n = 10)	Before TIPS placement	4–7 days after TIPS placement	1–2 months after TIPS placement
Values of octanoic acid breath tests (OBT) and aminopyrine breath tests (ABT) in 10 cirrhotic patients before and after TIPS placement. Data are given as median and (range). *P = 0.06 comparing OBT tmax values on the three time points. **P < 0.05 as compared with value before TIPS placement.
OBT t1/2 (min)	95 (71–142)	104 (72–138)	104 (75–207)
OBT tmax (min)	38 (21–78)	36 (29–72)	46 (33–94)*
OBTc (% dose)	50 (41–63)	49 (42–62)	52 (31–59)
ABT 2 h (% dose)	4 (1–8)	2 (1–3)**	2 (0–5)**

Discussion

Cirrhotic patients can develop delayed gastric emptying, independent of possible concomitant diabetes mellitus.³ The presence and/or endoscopic treatment of oesophageal and gastric varices might disturb gastric emptying rate, although reported data are conflicting.^3,4 Delayed gastric emptying can lead to malnutrition, which is a serious problem in the case of cirrhosis.^5,14 Gastric emptying rate is measured by Tc*-scintigraphy or, alternatively, by OBT, as has been validated and previously discussed.^1,2,16 Both examinations are considered standard methods for detection of delayed gastric emptying.¹⁶ The present OBT has the advantage over radioactive isotope studies in that it can be easily and safely repeated. Therefore, OBT was preferred in the present investigation.

The ‘postgastric processing’ of octanoic acid in OBT is measurable as ¹³CO₂ enrichment in the breath,^1,13 but the impact of liver function has not been studied before in adults.¹⁷ We quantified in cirrhotic patients the impaired degree of liver function by the Child–Pugh score¹⁴ and the microsomal liver breath test ABT.¹² Both parameters are most frequently used to assess the degree of cirrhosis.^18,19

At time point +∞, cumulative ¹³CO₂ recovery in the breath after oral octanoic acid intake was approximately 50% of the administered dose in healthy volunteers (Table 3).² Tracer recovery cannot be 100% because some metabolic steps and C₁-pooling do not fully lead to CO₂ expiration.⁹ In an ABT, ¹⁴CO₂ recovery after oral aminopyrine intake is much lower and approximates 20%.²⁰

In the first part of this study, we could not find a statistically significant difference between healthy controls and cirrhotic patients as to the parameters OBT t_1/2, OBT t_max or OBT_c. Apart from six of 82 cirrhotic patients who had a high OBT t_1/2 outside the range of healthy controls, data in both groups overlapped. Intriguingly, two correlations between OBT t_1/2 and liver function tests were observed (Table 4). First, OBT t_1/2 correlated weakly, although significantly, with aminopyrine demethylation. Secondly, OBT t_1/2 tended to correlate with the prothrombin time. These two parameters of hepatic microsomal and synthesis function, respectively, may be inversely related to a small delay in OBT t_1/2. Additionally a light change in the shape of the ¹³CO₂ expiration curves occurred in cirrhotic patients (Fig. 2). Yet, despite small OBT changes in cirrhotic patients, the Child–Pugh score, which combines five liver parameters,¹⁴ definitely did not correlate with OBT t_1/2. This indicates that liver function impairment as such did not delay OBT t_1/2, which mostly overlapped with healthy control data, as mentioned before. In clinical practice, this offers a likely argument, in our opinion, to ascribe a delayed OBT t_1/2 in a cirrhotic individual to delayed gastric emptying and not to his liver function impairment.

We acknowledge a limitation of this study, namely that a Tc*-scintigraphy was not performed together with an OBT²¹ in this large patient group. This would have corroborated our findings. On a whole, our observations suggest maintenance of octanoic acid oxidative breakdown in the case of severe liver impairment. At first, it was believed that, in cirrhotic patients, hepatic uptake and β-oxidation of (all) fatty acids increased.⁵ This was considered a necessary energy supply in a hypermetabolic state.⁵ Recently, normal²² or low²³ hepatic fatty acid extraction rates were observed in cirrhotic subjects. More methodological comparisons are certainly needed to clarify this problem of hepatic uptake ànd β-oxidation. An important element to take into account in this respect is the length of the carbon chain of the fatty acid. Hepatic β-oxidation of fatty acids in healthy men augment indeed with shorter carbon chains.^8,9 So was, after an oral gift, cumulative ¹³CO₂ breath recovery three times higher for 12 : 0 than 18 : 0 carbon fatty acids (41 ± 7 vs 13 ± 5%).⁹ Our data demonstrated a rather high recovery for octanoic acid in both healthy controls and cirrhotic patients. This suggests an extraction and oxidative breakdown of octanoic acid as efficient in cirrhosis as in healthy condition.

In the TIPS of the present study, we studied the influence of reduced portal vein supply to the liver. The portal vein is the main road of absorbed octanoic acid. During TIPS follow-up, there were no significant changes in OBT parameters, despite a worsening of aminopyrine demethylation. The OBT t_max tended to be later, 1–2 months after TIPS at a time point when none of the 10 patients had portal hypertension anymore (no patient required TIPS manipulations). More haemodynamic studies in cirrhotic and portal hypertensive patients are required to interpret possible changes in the shape of OBT curves. In general, because TIPS derives portal venous blood away from the liver²⁴ and OBT t_1/2 was similar after TIPS placement, this again may support the concept that the liver efficiently removes octanoic acid from the circulation.^6,11 Another possible explanation would be that extrahepatic metabolism of octanoic acid takes place or takes over after TIPS placement. Adipose tissue does not store medium-chain fatty acids (in contrast to long-chain fatty acids) but rather oxidizes them.⁶ More specifically, extrahepatic metabolism of octanoic acid has been described.¹¹ After intravenous injection of radioactive labelled octanoic acid to rodents, liver, kidney, heart and other organs were found to take up octanoic acid.¹¹ Hepatic and extrahepatic oxidative breakdown of octanoic acid are therefore not mutually exclusive. Whether small changes in the shape of OBT curves in cirrhotic patients are the result of extrahepatic oxidative breakdown is speculative and remains inconclusive in our study. Oral administration of octanoic acid in liquid form, as in our study, challenged hepatic uptake in the first place. Sufficient uptake and oxidative breakdown of octanoic acid by the liver and possibly other organs is an additional argument to use OBT as a gastric motility test in clinical practice. In our ten TIPS patients, our study had enough power to measure reduced microsomal liver function (ABT) after TIPS. These cirrhotic patients belonged to Child B and C classes – almost all were haemodynamically decompensated. Microsomal liver function can indeed worsen after TIPS placement as reported in decompensated cirrhotic patients²⁵ but does not need to worsen in patients with less advanced cirrhosis.²⁶

In conclusion, OBT t_1/2, an important parameter to quantify gastric emptying rate, was not changed by the presence and severity of cirrhosis. A TIPS derivation of portal perfusion of the liver did not affect octanoic acid oxidative breakdown despite diminution of microsomal liver function. The OBT may therefore be a suitable test in the future to detect delayed gastric emptying of solids in cirrhotic patients with reduced liver function and portal hypertension.

Acknowledgments

This study was supported by a grant of the Fund for Scientific Research (FWO Vlaanderen no 6.0111.98), Belgium, to FN.

The authors appreciate the contributions of G. Van Roey, E.A. El Atti, K.P. Geboes, B. Maes, A. Wilmer (Internal Medicine), G. Maleux and L. Stockx (Radiology), and N. Blanckaert and Z. Zaman (Clinical Laboratory), University Hospital Gasthuisberg, Leuven.