Jørgensen, Marianne H.*; Ott, Peter†; Michaelsen, Kim F.‡; Porsgaard, Trine§; Jensen, Fleming||; Lanng, Susanne*
Cystic fibrosis (CF) is the most common autosomal recessive disorder affecting the white population, at a frequency that varies between countries (1:2000–1:40,000 live births) (1). The most common CF mutation is F508del leading to severe clinical disease (2). This mutation accounts for 87% of mutated genes in the Danish population (3).
CF is a multiorgan disease, characterized by disturbances in electrolyte transport in exocrine glands and secretory epithelia. The main clinical features are recurrent pulmonary infections, decreased exocrine pancreatic function, intestinal obstruction, hepatobiliary disease, and salt loss.
Patients with CF have low levels of the n-3 long-chain polyunsaturated fatty acid (LCPUFA) docosahexaenoic acid (DHA) and a predominance of n-6 LCPUFA—especially arachidonic acid (AA)—in plasma and red blood cells (RBC) (4–11). These changes seem to be related to the severity of the mutation (10), although this may not be a consistent finding (12). It has been suggested that these differences are related to the severity of lung symptoms, and in this situation, supplementation with n-3 fatty acids as fish oil could provide some benefit for patients with CF, although larger interventions studies are needed.
In a healthy population, the nutritional intake of n-3 LCPUFA as fish or fish oil capsules is reflected in a dose-dependent way in the fatty acid profile in plasma and RBC membranes (13–15). Therefore, differences in LCPUFA status in patients with CF could be a result of differences in n-3 LCPUFA intake.
Mutations of the CF transmembrane conductance regulator protein (CFTR) are believed to play a major role in severity of CF disease (2). CFTR−/− knockout mice had lower levels of DHA and higher levels of AA compared with wild-type mice in organs positive for CFTR receptors (pancreas, ileum, and lung) (16). Interestingly, this was not seen in organs without CFTR receptors such as brain and kidney. Feeding CFTR−/− mice high levels of DHA as fish oil normalized fatty acid composition and resulted in normalization of anatomical changes in CFTR-positive tissues (16). In patients with CF, tissue differences have been shown in nasal biopsies, with lower levels of n-3 fatty acids and higher levels of n-6 fatty acids compared with healthy controls (17). Thus, findings from CFTR−/− knockout mice may also be present in patients with CF. Accordingly, differences of LCPUFAs in patients with CF may be more pronounced in tissues or compartments, which are influenced by CFTR receptors such as lymphocytes and mononuclear cells rather than plasma or RBC.
Chain elongation and desaturation of LCPUFA are mainly located in the liver (18). The enzyme system is shared by the n-3 fatty acids (eg, DHA) and n-6 fatty acids (eg, AA). In both chronic and acute liver disease without CF, n-3 LCPUFA in plasma and RBC are known to be lower than controls (19,20).
Approximately 30% of a CF population will develop liver disease mainly during the first and second decades of life (21,22). The diagnostic criteria of CF-related liver disease (CF-RL) are still debated and no specific diagnostic tools are available (23,24); patients develop fibrosis in the liver without elevated transaminases (25), hepatobiliary iminodiatic acid scanning and conventional ultrasound (25–27) are significant only at later stages of liver disease, and liver biopsies may underestimate disease (26,28). In many centres, CF-RL are pragmatically defined as either changes on ultrasound or hepatobiliary iminodiatic acid scan or elevated transaminases >1.5 times normal values measured at 2 occasions within 1 year without other explanations (eg, hepatotoxic drugs, severe infection) (22).
In the present study, we evaluated the fatty acid composition of RBC, granulocytes, and mononuclear cells in patients with CF and related them to the degree of liver disease measured by different methods.
In the Copenhagen CF cohort (n = 269), 66 patients fulfilled pragmatic criteria for CF-RL defined as otherwise unexplained elevated transaminases >40 U/L on 2 or more occasions within 1 year and/or pathological ultrasound of the liver. All 66 patients were treated with ursodeoxycholic acid (UCDA; 20 mg · kg−1 · day−1) started routinely after the second measurement of elevated alanine aminotransferase (ALT) or aspartate aminotransferase (AST) and all 66 patients have had elevated transaminases on 2 or more occasions. Twenty of the patients with CF-RL were matched with 20 patients with CF without known CF-RL defined as normal transaminases at routinely taken blood samples once per year during the last 10 years and no physical signs of liver disease. None of these were treated with UDCA. Matching criteria between the 2 groups were age, sex, body mass index, diabetes mellitus, P aeruginosa infection, and forced expiratory volume in 1 second (FEV1)%. Patients who had exogenous pancreas sufficiency and patients receiving lung transplantation were excluded.
To eliminate the influence of treatment with antibiotics on the transaminase level, the examination was performed at least 1 month after treatment with antibiotics. The examination included physical examination, blood samples and hepatic ultrasound, and hepatic transit time (HTT).
Blood samples were analysed for liver biochemistry and haematology. Pathological ALT and AST were set to be >40 U/L, whereas pathological γ-glutamyltransferase was set to be >70 U/L.
Granulocytes, mononuclear cells (pooled monocytes and lymphocytes), and RBC were separated by density gradient centrifugation using Histopaque 1119 and 1077 (Sigma Diagnostics, St Louis, MO) according to the manufacturer's procedure. Isolated granulocytes and mononuclear cells were washed in phosphate buffered saline (pH 7.2) and stored in methanol at −80°C until analysis. RBC membranes were burst with redistilled water before storage in methanol. Fatty acid composition of isolated cell fractions was determined after lipid extraction using chloroform/methanol (29), methylation with borontriflouride, and gas chromatography of the fatty acid methyl esters (30).
Ultrasound examinations of the liver were performed in all of the patients by the same investigator (F.J.), who was unaware of the UDCA treatment status of the patients. All of the patients were scored according to Williams’ ultrasound scoring scale (WUSS) (25). This includes description of abnormalities of the liver parenchyma, nodularity of the liver edge, and the presence of periportal fibrosis. Each variable were scored from 1 to 3.
HTT was measured with the ultrasound contrast agent SonoVue (Bracco Diagnostics Inc, Milan, Italy). The bolus consists of 20 μL of sulphur hexafluoride (SF6) microbubbles having a mean size of 3 μm. HTT is the time from an IV bolus injection in a cubital vein until arrival in a hepatic vein, detected with contrast harmonic imaging (Acuson, Sequoia, Mountain View, CA).
Pathological WUSS was set to be >3 points for all of the age groups. Pathological HTT was set to be <24 seconds for patients older than 18 years (31) because no reference values are available for children.
On the day of examination, weight, height, lung function as FEV1, and FVC (percentage of predicted values) were measured. All of the participants had a short questionnaire of frequencies of fish intake and regular intake of fish oil capsules.
All of the data were analyzed using SPSS 15.0 for Windows (SPSS Inc, Chicago, IL). Comparing difference of the mean, Student t test and 1-way analysis of variance were used. Frequencies were compared using χ2 or Fisher exact test if a frequency occurred <5 times. Linear regression (backward, with the fatty acid as the dependent variable) was used to test relations between parameters in a multivariate setting. The level of statistical significance was set to be P < 0.05.
Demographic parameters and liver biochemistry are shown in Table 1. The 2 groups were similar concerning distribution of age, body mass index, lung function, sex, and chronic infections (1 had P maltofilia and the remaining 15 had P aeruginosa). The genotype of the CFTR gene was known in 39 patients; 29 patients were homozygous for F508del, 10 were heterozygous for F508del, and in 1 patient no mutations were found. All of the patients were classified to have severe mutations (class III–IV mutations) (2) and all of the patients were substituted with pancreatic enzymes.
Fish or Fish Oil Intake
The questionnaire of frequency of fish intake was available in 39 of 40 patients. The results are shown in Table 2. Four patients had a regular daily intake of fish oil capsules. They had significantly higher amounts of EPA but not DHA in all 3 cell lines. Surprisingly, we did not find any relation between frequency of fish intake and any of the 3 fatty acids. Neither DHA, EPA, and AA nor AA/DHA ratio differed in any of the cell lines. This conclusion did not change if patients with fish oil intake were excluded from the analysis.
Because elevated transaminases were a criterion for liver disease, it was expected that the significantly higher ALT (P = 0.002), AST (P = 0.02), and γ-glutamyltransferase (P = 0.01) would be found within the CF-RL group (Table 1). In the control group, we found 4 patients who had ≥1 transaminases slightly above normal values: One had ALT >40 U/L (44 U/L), 2 had AST >40 (43 and 49 U/L), and 1 had both ALT and AST >40 U/L (43 and 72 U/L, respectively).
Despite almost normal bilirubin levels (range 3–34 μmol/L), we found higher total bilirubin among patients in the CF-RL group (P = 0.006). All patients with bilirubin >17 μmol/L belonged to the CF-RL group (n = 6). No differences were seen in either international normalized ratio or albumin between the 2 groups.
Fatty Acids Measurements According to Presence of CF-RL
Fatty acid compositions within monocytes, granulocytes, and RBC are shown in Table 3. No significant differences between groups were seen concerning n-6, n-3 LCPUFAs, or AA/DHA ratio in any of the cell types. The groups were defined by the pragmatic criteria for CF-RL, and even though they are internationally accepted, it is well known that they do not include all patients with liver disease.
When the CF-RL group and the controls were compared, WUSS scores were worse, with higher values in the CF-RL group (P = 0.02). As shown in Table 2, the agreement between the pragmatic criteria and the WUSS score was poor because 9 patients in the control group had pathological scores higher than 3, confirming that ultrasound must be added routinely to the evaluation of liver disease within patients with CF (24).
HTT was available in 28 of 40 patients because 12 (mean age 8.9 years; range 4.5–17.2 years) refused to have the extra needle prick that was necessary for contrast injection. Interpretations of these data are complicated by the lack of normal values for children, and in fact HTT and age tend to be correlated in our patients (r2 = 0.12; P = 0.07). Above the age of 18 years, the normal value is known to be 24 seconds. HTT data were available in 21 patients older than 18 years. Of these, 47% (n = 10) had pathological HTT <24 seconds, with no difference between the 2 patient groups.
To further explore the relation between LCPUFA and signs of liver disease, a multiple regression analysis was performed; HTT, age at examination, P aeruginosa infection, frequency of fish intake, diabetes mellitus, ± CF-RL, and WUSS were included in this analysis. DHA in mononuclear cell membranes was positively associated with HTT (P = 0.01; r2 (adjusted) = 0.14; Fig. 1) as the only statistically significant variable. The AA/DHA ratio within the mononuclear cells was negatively associated with HTT (P = 0.003; Fig. 2) and positively associated with WUSS (P = 0.03) (adjusted R = 0.37; P = 0.007). This indicates that pathologic liver tests will increase the n-6 fatty acid AA relatively to DHA. For RBC-LCPUFAs, no significant associations were seen. These findings indicate that in patients with CF, LCPUFA levels within white blood cells and not RBC are negatively associated with the degree of liver disease.
This is the first published study in which fatty acid profiles were assessed in patients with CF in RBC, mononuclear cells, and granulocytes, and an attempt has been made to relate them to the degree of liver disease. In a similar study, Freedman et al (32) reported that differences in AA/DHA levels between patients with CF and healthy controls were only present in lymphocytes and monocytes and not in neutrophils and RBC. They also found altered AA/DHA levels in nasal mucosa and the rectum of patients with CF as compared with healthy controls (17). In our study, the lower level of DHA and higher level of AA were found within the CF population and related to the degree of liver disease. Together, these studies indicate that analysis of fatty acid composition must be measured in CFTR receptor–positive cells if LCPUFA are studied within patients with CF.
The study was complicated by the fact that a criterion standard for the diagnosis of liver disease in CF does not exist. The pragmatic criteria are widely used as a compromise, but they have limitations (23,26). Thus, the pragmatic criteria may overlook liver disease without biochemical abnormalities, especially if ultrasound is not performed routinely (26), and patients with elevated liver enzymes due to drugs or infections may wrongly be classified as chronic liver disease. In the present study, the classification of patients according to these criteria was not related to the fatty acid patterns. At the same time, structural abnormalities in the liver as assessed by ultrasound (HTT and WUSS) were statistically significantly associated with higher levels of the n-6 LCPUFA AA and lower levels of the n-3 LCPUFA DHA in the fatty acid profile in mononuclear cells and granulocytes (Figs. 1 and 2). WUSS has been proposed as an objective measurement of the degree of LD in CF (25,26) and HTT as a measure of capillarization in patients with cirrhosis (31). Our data indicate that CF-RL could negatively affect the LCPUFA profile in these cell types. HTT could be a tool for investigating the degree of liver disease among patients with CF. Because signs of liver impairment mainly develop during first and second decades of life (17,28), reference values within the paediatric age group are needed.
We found a weak relation between n-3 LCPUFA status in the different cell lines and reported intake of fish oil capsules. There was no relation between fish intake and n-3 LCPUFA status, even though fish is the major nutritional contributor of n-3 fatty acid. This was unexpected because the same questionnaire differentiated DHA levels in lactating women (33). Because all of our patients had pancreatic insufficiency and thus expectedly insufficient LCPUFA uptake, we expected a stronger dose-response relation between nutritional intake of fish/fish oil capsules and n-3 content of the cell lines. Cell culture models indicate that the absence of the CFTR receptor rather than the presence of fatty acids determines the fatty acid profiles in the CF cell membrane (34). Our data may reflect this, if the findings of Andersson et al (34) can be confirmed in human studies.
The changes in LCPUFA with high n-6 and low n-3 fatty acids of patients with CF are of concern because elevated AA has been associated with altered inflammatory response (35). This could be of importance in relation to lung inflammation and recurrent infections in patients with CF, which is important for the long-term lung function outcome (36), although the clinical value remains to be proven. If LCPUFA have a significant role in progression of CF-related lung disease, our study raises concerns regarding the patients with CF-RL.
The relation between liver disease defined as abnormal HTT and fatty acid pattern was present in the mononuclear cells and granulocytes. In contrast to the red cells, the former cell types express the CFTR receptor. In animal studies, fatty acid abnormalities were more evident in tissues that expressed the CFTR receptor (16,17), and similar findings were reported in patients with CF compared with healthy controls (32). These studies are in agreement with our findings and support the hypothesis that the CFTR receptor may play a role in regulation of the fatty acid composition of the cell membrane (37).
In addition, our data suggest that liver disease per se could influence fatty acid patterns in tissues expressing the CFTR receptor. The liver plays a central role in fatty acid metabolism because it is the primary organ for chain elongation and desaturation of n-3 and n-6 fatty acids. Systemic disturbances in these metabolites most likely originate from disturbed hepatic metabolism. In the present study, a relation between liver disease and fatty acid composition was found for the CFTR receptor–positive mononuclear cells and granulocytes and not in red blood cells, an observation that remains to be fully understood.
Our study suggests that mononuclear cells and granulocytes may be a more sensitive compartment in which to measure such changes. We suggest that in further studies of LCPUFAs in patients with CF, fatty acid composition must be measured in these cell lines.
The authors thank Karen Jensen for technical assistance and are grateful for the support given by the Danish Association CF.
1. Dodge JA, Morison S, Lewis PA, et al. Cystic fibrosis in the United Kingdom, 1968–1988: incidence, population and survival. Paediatr Perinat Epidemiol 1993; 7:157–166.
2. Koch C, Cuppens H, Rainisio M, et al. European Epidemiologic Registry of Cystic Fibrosis (ERCF): comparison of major disease manifestations between patients with different classes of mutations. Pediatr Pulmonol 2001; 31:1–12.
3. Schwartz M, Brandt NJ, Koch C, et al. Genetic analysis of cystic fibrosis in Denmark. Implications for genetic counselling, carrier diagnosis and prenatal diagnosis. Acta Paediatr 1992; 81:522–526.
4. Lloyd-Still JD, Johnson SB, Holman RT. Essential fatty acid status and fluidity of plasma phospholipids in cystic fibrosis infants. Am J Clin Nutr 1991; 54:1029–1035.
5. Lloyd-Still JD, Bibus DM, Powers CA, et al. Essential fatty acid deficiency and predisposition to lung disease in cystic fibrosis. Acta Paediatr 1996; 85:1426–1432.
6. Henderson WR Jr, Astley SJ, McCready MM, et al. Oral absorption of omega-3 fatty acids in patients with cystic fibrosis who have pancreatic insufficiency and in healthy control subjects. J Pediatr 1994; 124:400–408.
7. Roulet M, Frascarolo P, Rappaz I, et al. Essential fatty acid deficiency in well nourished young cystic fibrosis patients. Eur J Pediatr 1997; 156:952–956.
8. Benabdeslam H, Garcia I, Bellon G, et al. Biochemical assessment of the nutritional status of cystic fibrosis patients treated with pancreatic enzyme extracts. Am J Clin Nutr 1998; 67:912–918.
9. Clandinin MT, Zuberbuhler P, Brown NE, et al. Fatty acid pool size in plasma lipoprotein fractions of cystic fibrosis patients. Am J Clin Nutr 1995; 62:1268–1275.
10. Strandvik B, Gronowitz E, Enlund F, et al. Essential fatty acid deficiency in relation to genotype in patients with cystic fibrosis. J Pediatr 2001; 139:650–655.
11. Coste TC, Deumer G, Reychler G, et al. Influence of pancreatic status and sex on polyunsaturated fatty acid profiles in cystic fibrosis. Clin Chem 2008; 54:388–395.
12. Cawood AL, Carroll MP, Wootton SA, et al. Is there a case for n-3 fatty acid supplementation in cystic fibrosis? Curr Opin Clin Nutr Metab Care 2005; 8:153–159.
13. Mozaffarian D, Psaty BM, Rimm EB, et al. Fish intake and risk of incident atrial fibrillation. Circulation 2004; 110:368–373.
14. Milte CM, Coates AM, Buckley JD, et al. Dose-dependent effects of docosahexaenoic acid-rich fish oil on erythrocyte docosahexaenoic acid and blood lipid levels. Br J Nutr 2008; 99:1083–1088.
15. Welch AA, Shakya-Shrestha S, Lentjes MA, et al. Dietary intake and status of n-3 polyunsaturated fatty acids in a population of fish-eating and non-fish-eating meat-eaters, vegetarians, and vegans and the product-precursor ratio [corrected] of alpha-linolenic acid to long-chain n-3 polyunsaturated fatty acids: results from the EPIC-Norfolk cohort. Am J Clin Nutr 2010; 92:1040–1051.
16. Freedman SD, Katz MH, Parker EM, et al. A membrane lipid imbalance plays a role in the phenotypic expression of cystic fibrosis in cftr(−/−) mice. Proc Natl Acad Sci U S A 1999; 96:13995–14000.
17. Freedman SD, Blanco PG, Zaman MM, et al. Association of cystic fibrosis with abnormalities in fatty acid metabolism. N Engl J Med 2004; 350:560–569.
18. Cabre E, Gassull MA. Polyunsaturated fatty acid deficiency in liver diseases: pathophysiological and clinical significance. Nutrition 1996; 12:542–548.
19. Clemmesen JO, Hoy CE, Jeppesen PB, et al. Plasma phospholipid fatty acid pattern in severe liver disease. J Hepatol 2000; 32:481–487.
20. Socha P, Koletzko B, Pawlowska J, et al. Essential fatty acid status in children with cholestasis, in relation to serum bilirubin concentration. J Pediatr 1997; 131:700–706.
21. Scott-Jupp R, Lama M, Tanner MS. Prevalence of liver disease in cystic fibrosis. Arch Dis Child 1991; 66:698–701.
22. Colombo C, Battezzati PM, Crosignani A, et al. Liver disease in cystic fibrosis: a prospective study on incidence, risk factors, and outcome. Hepatology 2002; 36:1374–1382.
23. Colombo C, Battezzati PM, Strazzabosco M, et al. Liver and biliary problems in cystic fibrosis. Semin Liver Dis 1998; 18:227–235.
24. European Association for the Study of the Liver. EASL Clinical Practice Guidelines: management of cholestatic liver diseases. J Hepatol 2009;51:237–67.
25. Williams SG, Evanson JE, Barrett N, et al. An ultrasound scoring system for the diagnosis of liver disease in cystic fibrosis. J Hepatol 1995; 22:513–521.
26. Williams SM, Goodman R, Thomson A, et al. Ultrasound evaluation of liver disease in cystic fibrosis as part of an annual assessment clinic: a 9-year review. Clin Radiol 2002; 57:365–370.
27. Colombo C, Crosignani A, Battezzati PM, et al. Delayed intestinal visualization at hepatobiliary scintigraphy is associated with response to long-term treatment with ursodeoxycholic acid in patients with cystic fibrosis-associated liver disease. J Hepatol 1999; 31:672–677.
28. Lewindon PJ, Shepherd RW, Walsh MJ, et al. Importance of hepatic fibrosis in cystic fibrosis and the predictive value of liver biopsy 2. Hepatology 2011; 53:193–201.
29. Folch J, Lees M, Sloane Stanley GH. A simple method for the isolation and purification of total lipides from animal tissues. J Biol Chem 1957; 226:497–509.
30. Porsgaard T, Xu X, Gottsche J, et al. Differences in the intramolecular structure of structured oils do not affect pancreatic lipase activity in vitro or the absorption by rats of (n-3) fatty acids. J Nutr 2005; 135:1705–1711.
31. Albrecht T, Blomley MJ, Cosgrove DO, et al. Non-invasive diagnosis of hepatic cirrhosis by transit-time analysis of an ultrasound contrast agent. Lancet 1999; 353:1579–1583.
32. Freedman SD, Sullivan BPO, Martinez-Clark P, et al. An abnormality in AA/DHA metabolism is present in cftr-regulated tissues from CF patients. Pediatr Pulmonol 2000;20:158–9.
33. Jørgensen MH, Hernell O, Hughes E, et al. Is there a relation between docosahexaenoic acid concentration in mothers’ milk and visual development in term infants? J Pediatr Gastroenterol Nutr 2001; 32:293–296.
34. Andersson C, Al-Turkmani MR, Savaille JE, et al. Cell culture models demonstrate that CFTR dysfunction leads to defective fatty acid composition and metabolism. J Lipid Res 2008; 49:1692–1700.
35. Calder PC. n-3 polyunsaturated fatty acids, inflammation, and inflammatory diseases. Am J Clin Nutr 2006; 83:1505S–1519S.
36. Courtney JM, Ennis M, Elborn JS. Cytokines and inflammatory mediators in cystic fibrosis. J Cyst Fibros 2004; 3:223–231.
37. Carlstedt-Duke J, Bronnegard M, Strandvik B. Pathological regulation of arachidonic acid release in cystic fibrosis: the putative basic defect. Proc Natl Acad Sci U S A 1986; 83:9202–9206.