Breath test using ¹³C methacetin does not seem to be useful in the assessment of liver function in girls with anorexia nervosa: a case control study

Anorexia nervosa (AN) concerns approximately 0.3–4% of women, including also girls below 18 years old (around 0.8% to 1.8% of the pediatric female population) [1]. Hypetransaminasemia in the course of AN may occur as well in the period of decreased energy intake as also during realimentation [2]. Several studies in AN patients found an inverse correlation between the liver enzymes’ activity and BMI value, what suggests an important role of the nutritional status, but the mechanism of this phenomenon has not been clearly explained [3, 4]. Among possible causes cellular apoptosis and hepatocyte autophagocytosis induced by chronic malnutrition is considered. It seems probable that in the first phase of the disease, related to gradual weight loss, a moderate elevation of the activity of liver enzymes represents an attempt of the body to adapt to the decreased energy supply and protects the cells from premature death [2]. In most cases, hepatopathy in individuals with AN is mild and transient, and acute liver failure in AN is quite rare. Mild hypertransaminasemia (< 200 IU/L) concerns 30% to even 75% of AN patients [2, 5]. Usually the activity of aminotransferases exceeds the upper normal limit 2 to 4 fold, but significantly higher values can be also present [6]. As starvation continues and BMI falls below a critical value (in adults ≤13 kg/m²) an excessive activation of the mentioned above processes leads to destruction of hepatocytes [2]. There are few case reports of acute liver insufficiency episodes in these patients, but the total prevalence of this phenomenon is unknown [7, 8].

Another critical moment is the period of nutritional therapy with the risk of the so called refeeding syndrome (RS) as a response to the increased calories supply. Its complications cover disorders of the cardiovascular and nervous system as well as water-electrolyte imbalance and death, that may concern even 6% of adolescents hospitalized due to AN. [2]

Because of the above mentioned causes various guidelines and recommendations, including those of the American Psychiatric Association, advise constant supervision of laboratory parameters in patients with eating disorders, including even daily liver enzyme testing [9‐12]. Ridout and coauthors revealed however, that such a proceeding does not translate into therapeutic decisions and generates at the same time costs and burden disproportionate to the obtained benefits [13].

Presently breath tests are increasingly applied for the assessment of the function of various organs. Substances such as ¹³C methacetin are metabolized after oral administration by the cytochrome P450. During their decomposition ¹³CO₂ is formed and the rate of its excretion indicates the microsomal function of the liver [14]. While biochemical parameters show only hepatocyte injury and provide indirectly information about the organ function, the breath test results allow to estimate the degree of organ damage and also determine quantitatively the functional mass of the organ. Compared to the liver biopsy, which remains the gold standard of hepatological diagnostic, breath tests help to estimate not only the organ’s functional reserve but also to assess the long-term prognosis [15]. They are repeatable, easy to perform and noninvasive, what seems to be especially important in pediatric patients. However present experience concerning this method is mainly based on adult studies and only few investigated it in children. Most of the trials concerned patients with chronic, primary liver diseases or obesity [16]. Until now, to our knowledge, only results of one study on the usefulness of the breath test in the diagnostics of girls with AN were published, but the analysis was performed in a relatively small group of patients [17]. In addition this test seems to have important limitations in case of AN patients which are related to the specifics of the disease itself as well as its complications. For the above reasons this study was designed to analyze the utility of the noninvasive breath test using ¹³C carbon isotope labeled methacetin (Methacetin Breath Test - MBT) in the assessment of the liver function in girls with eating disorders of AN-type.

Hypertransaminasemia is one of the most common liver phenomenon observed in AN patients [6]. Within our study group hypertransaminasemia was revealed in 12.2% females with AN, wherein accordingly to data published by other authors, slightly higher ALT than AST activity was observed (24.99 IU/L ± 22.13 and 23.99 ± 14.11 IU/L respectively). None of the studied patients presented clinical or biochemical signs of liver insufficiency and the highest values of the liver enzymes activity did not exceed more than 3 fold the upper normal limit.

Application of the MBT in the assessment of the liver function is a promising diagnostic option, which is especially interesting in pediatrics, particularly in groups requiring constant or repeated monitoring of this organ’s function. MBT is indisputably one of the best described metabolic breath tests. Although the efficiency of MBT in the diagnostics of liver function was confirmed in different diseases, also in children, many factors may negatively influence the investigative value of this test. In the case of AN already the exactness of the execution of the test’s methodology may rise doubts. The test’s protocol assumes efficient consumption of 75 mg of ¹³C methacetin (odorless powder) dissolved in 200 ml of neutral, room-temperature fluid. Taking into consideration the psychopathology of the illness, ingestion of such amount of fluid may be difficult or even traumatizing for the patient. In two of our patients vomiting occurred soon after the ingestion of the substrate and test had to be discontinued. One girl could not consume the substrate because of a hysteria episode. Another limitation may be the duration of the test. During the 120 min test time the patients must remain fasting what may be in opposition to the refeeding plan whereas, in this group the nutritional refeeding should be priority.

Apart from psychological limitations also gastroenterological aspects of AN may play an important role. In AN patients such disorders as regurgitations, dysphagia, odynophagia, nausea or gastro-esophageal reflux can be observed [19]. These symptoms may lead to difficulties in the intake of ¹³C methacetin or ingestion of an inadequate amount of the substrate. Moreover AN is related to disorders of the motility of the gastrointestinal tract, including delayed gastric emptying and prolonged gut passage [19]. In our studied group of AN patients TTP was longer compared to normal ranges given by the manufacturer, what may be caused by the delayed passage and – in consequence – postponed absorption of the substrate. In the study group the maximal mean value of dose/h was in the 20th minute of the test. The time from ¹³C methacetin administration to the peak excretion of ¹³CO₂ in the exhaled air was usually between the 20th and 40th minute of the test (normal range: 10th–20th minute of the test). A similar situation was revealed also in the control group. It needs to be however underlined that the presumed normal ranges refer to measurements in adults, because there are no separate normal/reference values for the pediatric population. This fact might influence the present observations. Piwczyńska and coauthors, who investigated MBT utility in a group of children with autoimmune hepatitis, described significantly longer TTP in patients with severe liver fibrosis. The authors’ explanation of these results was the loss of metabolically active liver tissue during fibrosis [20]. Similar observations were described by Shteyer and coauthors who found significantly longer TTP in infants with biliary atresia. According to the authors it reflected the decreased activity of the hepatic P450 cytochrome in this group [21]. In our patients with AN, despite prolonged TTP, the mean maximal percentage for dose/h which describes the recovery of the substrate, was high - significantly higher than in the control group (31.29 ± 7.62% vs 24.61 ± 6.59%; p = 0.000). This finding may suggest rather stimulation of the hepatic methacetin metabolism and a different than liver insufficiency cause of delayed TTP.

In an earlier study that assessed liver function in girls with AN by means of MBT significant acceleration of the liver ¹³C methacetin metabolism was described when compared to the control group and the differences in %CD values between children with AN as controls were significant for all measurements after the 30th minute of the test. The authors’ interpretation of the obtained data was that girls with AN have a specific model of ¹³C methacetin metabolism which is related to, among others, increased recovery of the cumulative dose of the substrate. The mentioned study was however performed in a small group of children (n = 25, only 12 with AN) what importantly limited the reasoning power [17]. Our investigation of over 80 females confirms the significantly higher ¹³CO₂ cumulative dose values in the exhaled air in the study group in comparison to controls. Statistically significant differences concerned in this case all measurement time points. The dissimilarity of the methacetin metabolism in the group of AN patients could on one side evidence an actual increase of the liver’s cytochromal activity, but on the other side it may reflect the influence of exogenous factors on the measurement results. Existing studies on the impact of malnutrition on the activity of the liver P450 cytochrome were conducted in various age groups using diverse methodology and, in consequence, present ambiguous results. Hamon-Wilcot and coauthors investigated the activity of the CYP1A2 subunit of the P450 cytochrome in individuals above the age of 65 years with weight loss and hypoalbuminemia. The activity of the cytochrome was assessed using the paraxanthine/caffeine metabolic index. Its values in the study and control group were similar. There were also no correlations between BMI and the CYP 1A2 subunit activity [22]. On the other hand Oshikoya and coauthors by means of a ¹³C caffein breath test (CBT) proved significantly decreased recovery of the cumulative dose of the substrate (%cum dose) in malnourished children in the course of marazmus and kwashikor in comparison to the test’s results that were obtained after realimentation [23]. The outcomes of the last of the mentioned studies seem to stay in opposition to our observations. Nonetheless it needs to be pointed out, that in contrast to methacetin, caffeine is a substance characterized by a low hepatic extraction index and the CBT is considered to be relatively resistant to exogenous factors [24]. Therefore it seems to be essential to separate the impact of weight and nutritional status from other factors related to AN that might influence the MBT result. Studies analyzing the relations between the liver’s functional capacity and the left ventricle ejection fraction (LV EF) as well as the dimensions of the left heart showed a correlation between %CD of ¹³C methacetin in the 120th minute of the test and the stage of heart insufficiency (according to NYHA). The value of the ¹³CO₂ CD was also significantly lower in all time points in patients with lower LV EF. %CD ¹³CO₂ weekly correlated with LV EF as well as it had a week invert association with the LVEDd [25]. Our studied AN patients did not reveal signs of heart insufficiency. Their mean LV EF was 69.14 ± 5.092% and mean LVEDd – 4.202 ± 0.355 cm. This however does not rule out the possible impact of a single LV EF value on the obtained %CD¹³CO₂. The observed acceleration of the hepatic methacetin metabolism in the study group may also be a consequence of the patients’ medication. At the same time the increased liver metabolism of the CYP1A subunit could in a difficult to predict way modify the pharmacokinetics of the drugs used in AN treatment.

Undoubtedly an important role in the interpretation of the obtained results play the selection criteria for the study group. Tsukamoto and coauthors investigated 25 adult patients with eating disorders before realimentation was started. Elevated ALT activity was found in 13 of them and these were patients with confirmed AN. Individuals with incorrect ALT values had lower BMI and shorter disease duration. The BMI value was considered to be a significant and independent determinant in the process of liver injury in patients with AN [26]. Similar observations were conducted by Hanachi and coauthors, who were looking for risk factors of hypertransaminasemia in a cohort of AN patients obtaining nutritional therapy [3]. In our study we used the BMI percentile values in order to compare children of different age. The mean BMI percentile in the study group was 7.959 ± 11.696, with the values varying from the 50th below the 1st percentile for age. This wide range of this parameter results from the fact, that the study group included also girls with so called AAN (who fulfill the B and C criteria of the DSM-5 classification accompanied by a significant weight loss, but whose present weight was within normal ranges for age and sex). Patients with AAN have typically higher weight (at the beginning of the disease even obesity) and more intensive nutritional restrictions compared to typical AN, but the risk of organ complications is in both cases is similar [4]. On one hand it can be concluded, that higher BMI in girls with AAN, being an indirect evidence of better nutritional status, is related more to adaptive changes of the liver metabolism and to the domination of cytoprotective over cytolytic processes leading to the preservation of the normal liver function. However on the other hand females with AAN usually undertake more intensified actions aimed at fast reduction of the weight, what combined with the typical for adolescents high basic metabolism index, may lead to earlier development of metabolic complications [3]. This group may therefore have a difficult to foresee liver metabolism status which influence the general outcomes of our study. The multiplicity of the presented above factors that may potentially interfere with the MBT results casts in doubt the usefulness of this method in patients with eating disorders of AN type.

The practical implication of our finding is that described by us peculiar feature of MBT in AN patients may decrease the test’s sensitivity in this group. The percentage of the exhaled ¹³CO₂ cumulative dose in the 120 min of the test is one of the basic parameters in result interpretation. Higher %CD 120, due to increased liver cytochromal activity, result in normal test outcomes and create a risk of false negative results. A potentially higher percentage of false negative results significantly decreases the diagnostic value of this method. Nevertheless, this study has also some limitations that require mentioning. The study group included only mild hypertransaminasemia patients and therefore it is difficult to predict if the findings would be similar in AN individuals with severe hypertransaminasemia and/or liver failure. The background of our outcomes is still unclear. Two factors that may influence the MBT results in AN patients need to be considered. Firstly, it seems to be critical to estimate if the test results depend on weight and nutrition status of the patient. Secondly it is crucial to exclude other than undernutrition factors that can impact the test’s results. A prospective study would be helpful to analyze whether the methacetin metabolism changes after realimentation.

The strengths of the study include its case-control design that allowed to compare MBT parameters in AN patients with sex and age-matched children. We also excluded heart failure in patients from the study group - one of the most important factors that influence the MBT results. Furthermore to the best of our knowledge this is the first study that estimates the utility of ¹³C MBT in a representative group of patients with eating disorders of AN type.

The obtained results confirm that the liver ¹³C methacetin metabolism in female patients with AN is significantly increased in comparison to healthy controls. These observations, apart from clinical implications related to the possible increased activity of the enzymatic system of the liver P450 cytochrome, generate also a question concerning the usefulness of the MBT in the assessment of the liver function in this specific group of patients. This method still raises many interpretational doubts. Increased cumulative dose recovery of the substrate in girls with AN influences the reliability of the result. The described phenomenon creates also a risk of obtaining a higher percentage of false negative results, what significantly decreases the diagnostic value of this method. There is a need for further clinical studies that would assess the influence of exogenous factors related to AN on the test’s result.