Current Treatment Options for Cystic Fibrosis-Related Liver Disease

Cystic Fibrosis-related liver disease (CFLD) has become a leading cause of morbidity and mortality in patients with Cystic Fibrosis (CF), and affects children and adults. The understanding of the pathogenesis of CFLD is key in order to develop efficacious treatments. However, it remains complex, and has not been clarified to the last. The search for a drug might be additionally complicated due to the diverse clinical picture and lack of a unified definition of CFLD. Although ursodeoxycholic acid has been used for decades, its efficacy in CFLD is controversial, and the potential of Cystic Fibrosis Transmembrane Conductance Regulator (CFTR) modulators and targeted gene therapy in CFLD needs to be defined in the near future. This review focuses on the current knowledge on treatment strategies for CFLD based on pathomechanistic viewpoints.


Pathogenesis of CFLD
The pathogenesis of CFLD is believed to be multifactorial and not clarified to the last [17]. Thinking of the whole clinical picture of CFLD that could be present, i.e., cholangiopathy, steatosis, liver fibrosis, focal cirrhosis, and portal hypertension (without cirrhosis being a mandatory requirement), we probably need to distinguish between primary, secondary, and tertiary causes that also might influence each other.
Primary causes for liver alterations, such as genetic factors affecting bile acid homeostasis and epithelial innate immunity of the bile ducts, secondary causes such as CF-related diabetes (CFRD), exocrine pancreatic insufficiency, intestinal inflammation, changes in the gut microbiome, as well as tertiary causes such as antibiotics, immunosuppressive medication, and high fat diet might coin the clinical presentation of CFLD.
CFTR mutations are classified as class I, II, III, IV, V, and VI depending on the underlying CFTR defect. Class I mutations affect biosynthesis of CFTR and thereby lead to the absence of a functional CFTR protein. Class II mutations (including the most common CFTR mutation deletion of a phenylalanine at position 508 (F508del)) interfere with protein maturation, why CFTR comes up misfolded and causes a CFTR trafficking defect. Class III mutations lead to a defective Cl-/HCO3 − -channel regulation. Class IV mutations cause a decreased Cl − /HCO3 − -channel conductance, Class V mutations cause a reduced synthesis of CFTR, and Class VI mutations lead to CFTR membrane stability [2].
In the liver, CFTR is expressed in the apical membrane of cholangiocytes, but not hepatocytes [19], and constitutes a major determinant of bile secretion. Dysfunctional or lacking CFTR protein leads to a disrupted Cl − -and HCO3 − secretion, and thus a hyperviscous and more acidic bile (including aberrant bile acid composition and increased concentration of toxic bile acids). The resulting ductular biliary obstruction leads to local inflammation resulting in the picture of sclerosing cholangitis as well as portal inflammation, which may cause a focal fibrosis in the first place, but later multilobular fibrosis and cirrhosis. Furthermore, as recently described by Fiorotto et al., CFTR is additionally involved in the control of biliary epithelial inflammation and permeability in a mouse model [20]. This is facilitated by controlling the activation of the tyrosine kinase Src, a protooncogene, which in turn regulates toll-like receptor 4 (TLR-4)-responses to gut-derived endotoxins. Additionally, aberrant TLR-4 activation decreases the epithelial barrier function leading to a back-diffusion of toxic bile acids again causing peribiliary inflammation and fibrosis [20]. Additionally, the same researchers recently postulated an auto-inflammatory component in the pathogenesis of CFLD as a result of aberrant activation of the innate immune system. This may be mediated via an increased basal NF-kB activation found in human F508del cholangiocytes, which in turn actively secrete pro-inflammatory chemokines known to attract neutrophils. Furthermore, F508del cholangiocytes showed increased responsiveness to endotoxins, as well as Src kinase activation and toll-like receptor 4 (TLR4) phosphorylation leading to local inflammation [21].
A clear definition of CFLD, as well as unraveling the pathomechanism that underlies the evolution of CFLD, including the consideration of its clinical characteristics and severity determine the development of efficacious treatments.

General Treatment Recommendations for CFLD
Treatment of CFLD should be in the hand of an experienced, multidisciplinary team [25]. It comprises two major aspects, firstly, treating the underlying liver disease itself, and secondly, managing nutritional therapy, as well as portal hypertension and decompensated cirrhosis, both possible results of the underlying liver disease. Liver transplantation may be the ultimate treatment in patients with end-stage liver disease.
Treating the underlying liver disease to date mainly includes the administration of ursodeoxycholic acid (UDCA). Current knowledge on UDCA in CFLD is described in Section 5. The potential of CFTR modulators, gene therapy and future treatment options is described in Sections 6-8.
Nutritional therapy with regard to CFLD mainly aims at improving the nutritional status of the patient by optimizing dietary intake of calories (especially protein), pancreatic enzyme replacement, supplementation of fat-soluble vitamins A, D, E, K, and insulin treatment in patients with CFRD. It may also include the correction of deficiencies of essential fatty acids, carnitine and choline since an association with liver steatosis has been postulated [25,26].
The management of portal hypertension in CFLD (especially adulthood CFLD) includes endoscopic screening for gastroesophageal varices and consecutive band ligation or sclerotherapy as primary or secondary bleeding prophylaxis [27]. Unselective beta-blocker therapy with propranolol or carvedilol might be used in order to prevent bleeding [27], but should be used with caution in patients with reactive airway disease, since randomized controlled trials in patients with CF are currently lacking. Other complications of portal hypertension such as ascites, spontaneous bacterial peritonitis and hepatic encephalopathy should be treated according to current guidelines on patients with decompensated liver cirrhosis [28]. However, it needs to be taken into account that portal hypertension in patients with CFLD often occurs in the absence of liver cirrhosis and the presence of preserved liver function [29]. Non-cirrhotic portal hypertension in CFLD was found to be associated with portal branch venopathy in small case series [30,31]. In these patients, transjugular intrahepatic portosystemic shut can be a therapeutic option.
Unified criteria for the indication and timing of liver transplantation for CFLD have not been established so far. However, in patients with CF, deterioration of nutritional status and pulmonary function should be given special consideration [10]. Since the median survival time has been increasing, liver transplantation for CFLD might become more frequent in the future.

Effects of Ursodeoxycholic Acid on CFLD
As according to the Best Practice Guidance for the Diagnosis and Management of CFLD, UDCA should be commenced as soon as the diagnosis of CFLD has been made at a dose of 20 mg/kg body weight (BW) in order "to delay the progression of the disease" [10].
However, although UDCA has been repeatedly investigated as treatment for CFLD since the 1990s [32,33], until now its efficacy in CFLD has remained controversial. Several small prospective case series and larger retrospective registry data mainly could demonstrate a significant positive effect on liver enzymes, however, randomized controlled trials assessing hard endpoints such as improvement in liver histology, mortality or liver transplant free survival are still lacking but remain crucial. Furthermore, the interpretation of the findings is frequently limited due to the lack of a homogenous definition of CFLD and uniform assessment of liver disease severity. Table 1 summarizes the current knowledge on UDCA in CFLD-treatment.
The lacking evidence is made obvious by a Chochrane review from 2017, which analyzed the present evidence on UDCA in terms of improvement of liver function, risk reduction for the development of chronic liver disease, and improvement in general outcomes in CF [34]. All randomized controlled trials that applied UDCA for at least three months compared with placebo or no additional treatment were considered. Of 12 identified trials only three trials comprising 118 participants dating back to the 1990s were included [35,36], applying 10-20 mg/kg body weight of UDCA for up to 12 months. No significant effect of UDCA could be found, not at least also due to the fact that patient numbers analyzed were very small, and length of treatment with UDCA might have been too short. A more recent longitudinal cohort study by Toledano et al. (already mentioned above) investigating >3000 patients reported a positive association of UDCA with prolonged overall survival in patients without cirrhosis (HR 0.50, 95% CI 0.36-0.69, p < 0.0001) in contrast to patients with cirrhotic disease (HR 1.19, 95% CI 0.46-3.10, p = 0.71) [16], suggesting a positive effect in patients with mild CFLD.
In summary, although the use of UDCA in CFLD has become a standard treatment, the scientific basis for this requires more reliable data especially for efficacy endpoints in middle-and long-term use. Nevertheless, UDCA appears to be beneficial, especially when started early and against the background of lacking therapeutic alternatives.

Effects of CFTR Modulators on CFLD
CFTR modulators are small molecules, which directly target CFTR and thereby partly restore altered CFTR protein function. Their development has presented a breakthrough in the treatment of CF, however, their effect on CFLD remains widely unclear.
Generally, we need to distinguish between CFTR potentiators and correctors, as well as stabilizers, amplifiers and read-through agents, which may be given in supplement to the first two. Potentiators keep the CFTR channel open so that chloride transport is ensured, correctors may help to ensure correct CFTR protein folding so that CFTR is able to traffic to the cell surface and remains there longer. In addition, stabilizers promote CFTR maturation and plasma membrane stability, and thereby increase the half-life of the CFTR protein, amplifiers may increase the amount of CFTR protein that is produced by the cell, and read-through agents may bypass premature termination codons. Currently, only CFTR potentiators and correctors are readily available to patients.
The CFTR modulator which was developed first was ivacaftor (VX-770, Kalydeco ® ), a CFTR potentiator. The small molecule treating the class III mutation G551D gained FDA approval in 2012. Later it became apparent that it also works in other class III and IV mutations (gating, residual function, splice and conduction mutations). The second medication which was approved by the FDA in 2015 was a CFTR modulator combination of ivacaftor with lumacaftor (VX-770/VX-809, Orkambi ® ) to treat homozygous F508del mutation (protein processing mutation). However, its effect on improvement of FEV1 and reduction of pulmonary exacerbations was only small to modest, probably also due to antagonistic effects of VX-809 and VX-770. In February 2018 tezacaftor, another corrector, was approved in combination with ivacaftor (VX-661/VX-770, Symdeko ® ) for patients with at least one F508del mutation (protein processing, residual function, splice mutations), bearing the advantage of lesser side effects and drug interactions than the combination of lumafactor/ivacaftor. Ivacaftor, lumacaftor, and tezacaftor all are deemed as first generation CFTR modulators. Although their development represented a breakthrough in the management of patients with CF, treatment with the first generation CFTR modulators either cannot target frequent mutations, or their efficacy is limited. This has highlighted the need for further drug evolution and led to the development of the next generation CFTR corrector VX-445/ elexacaftor, which is able to correct an additional defect in the formation of the F508del-CFTR protein. Hence, a drug was approved as a triple combination of ivacaftor, tezacaftor, and elexacaftor (VX-770/Vx-661/VX-445, Trikafta™) by the FDA in October 2019. Trikafta™ is approved for patients with at least one F508del mutation and, in contrast to its predecessors, shows higher efficacy and is estimated to be able to treat about 90% of the CF population [50][51][52]. Table 2 presents an overview on approved CFTR modulators and their effects.
Further CFTR modulators currently are being tested in phase 1 or 2, such as the CFTR potentiators VX-561 (deuterated ivacaftor) and ABBV-3067, the CFTR correctors ABBV-2222, and VX-121, the CFTR amplifiers fPTI-428, and PTI-CH, and the read-through agent ELX-02 [53]. -normalization of ALT, AST and GGT in all pts -no effect on focal biliary cirrhosis -at least two of the following conditions were present: hepatomegaly confirmed by US, other abnormalities of the liver parenchyma on ultrasound, e.g., heterogeneous echogenicity and persistently increased liver enzymes (ASAT, ALAT, GGT) with at least two out of these three being abnormal for at least 12 mo -persistently raised transaminases only -US examination consistent with cirrhosis, i.e., nodular aspect of the liver and/or a clearly irregular liver contour and/or splenomegaly 15-20 mg 103.2 n.r.
-decrease of ALT (p < 0.001), AST (p < 0.01), and GGT (p = 0.01) -Decrease in liver stiffness (p < 0.01) in a subgroup of patients For all currently available CFTR modulators in clinics it is recommended to check for liver function test abnormalities (transaminases and bilirubin) on a regular basis, which have been described in about 5% to 15% of patients as a potential side effect of these drugs up to >8 time the upper limit of normal [50,54,55]. However, there is also some evidence that CFTR modulators might have a beneficial effect on the liver. Van de Peppel et al. described that treatment with Ivacaftor partially restored disrupted FGF19-regulated bile acid homeostasis in 117 patients with CFTR gating mutations (partially F508del heterozygoty) participating in the GOAL study, yet these findings did not correlate with CFTR function in other organs, as measured by sweat chloride levels or pulmonary function [56]. Furthermore, mid-and long-term effects on liver function or histology remain unclear. Another study by Gelzo and colleagues reported a beneficial effect of Lumacaftor/Ivacaftor on cholesterol metabolism, enterohepatic flux and improvement of alkaline phosphatase in 40 patients with at least one F508del mutation [57]. A small study including 20 patients with CF, of which nine received Lumacaftor/Ivacaftor, the use of this CFTR modulator combination was associated with less hepatic steatosis as assessed by magnetic resonance imaging proton density fat fraction [58]. In an in vitro study with human derived pluripotent stem cells induced to cholangiocytes, Fiorotto et al. could show that the effect of Lumacaftor/Ivacaftor to correct and potentiate F508del CFTR by a combination with the Src inhibitor PP2 successfully restored fluid secretion to normal levels [21].
Ivacaftor was also hypothesized to be effective in restoring defective phosphatidylcholine secretion in five mutations in the ATP-binding sites of ABCB4, three of which five have been identified as gating mutations in CF (G551D, S1251N, and G1349D) in cell models supplemented by a three-dimensional structural modeling [59].
Finally, Ivacaftor has also shown therapeutic potential in an in vitro model of progressive familial intrahepatic cholestasis type 2 (PFIC2) caused by ABCB11 missense mutations affecting bile salt export pump (BSEP): Ivacaftor treatment increased the taurocholate transport activity of mutated BSEP by 1.7-fold, reaching 95% of BSEP function [60].  Increased liver enzymes (8%) and bilirubin (3%) > 2x ULN Significant increase in FEV1% through day 29 [50] restores FGF19 regulated bile acid homeostasis [56] increase in small intestinal pH [61] increase in Akkermansia, decrease in intestinal inflammation [62] ivacaftor  [52] In summary, CFTR modulators show a certain therapeutic potential in CFLD, but there is still a long way to go and we are just at the beginning. We need to learn more about the mechanisms underlying hepatic toxicity as well as improvement of liver alterations by CFTR modulators in order to select patients who may benefit. This evolution might be inhibited by the fact that there is no clear association of CFLD with genetic mutations, and that a unified definition of CFLD based on its pathomechanism and pathophysiology is currently lacking. Thus, to date the therapeutic response to CFTR modulators seems unpredictable.

Effects of Gene Therapies on CFLD
In contrast to CFTR modulators, which partially restore CFTR protein function, CFTR targeting gene therapies establish a new, correct version of the CFTR gene in order to produce normally functioning CFTR protein. In contrast to CFTR modulators, gene therapy bears the advantage that patients may be treated independently of the underlying CFTR mutation.
A randomized controlled trial (phase 2b) investigated a nebulized pGM169/GL67A gene-liposome complex delivering plasmid DNA encoding the CFTR gene to the lungs and demonstrated a significant, but modest effect on FEV1% after 1 year of monthly application [66]. Notably, this non-integrating gene therapy had no negative effect on liver enzymes. However, the improvement of lung function was lower in comparison to those achieved with CFTR modulators, so that the substance did not reach clinical application.
Another approach is to use messenger ribonucleic acid (mRNA)-based treatments, such as MRT-5005, to deliver the correct mRNA to lung epithelial cells in order to produce functional CFTR protein. Currently, a clinical phase1/2 trial (NCT03375047, RESTORE-CF) is underway.
Further possibilities evolve by the use of clustered regulatory interspaced short palindrome repeats (CRISPR)/Cas9 to repair defective CFTR in vitro and in vivo [67][68][69], but have not been tested in models of CFLD or humans so far.

Conclusions and Future Treatment Options
There has been tremendous effort and progress in the treatment of CF within the last years. Due to the favorable epidemiologic evolution we have to face increasing numbers of patients suffering from CFLD. However, to date, an efficient causative therapy is lacking.
Besides the evolution of CFTR modulators restoring CFTR protein function, substances creating an additive effect, such as Src inhibitors targeting the TLR-4-mediated inflammatory processes at the bile duct epithelium could be a future treatment option, but have not been investigated in humans. Furthermore, molecules involved in CFTR ubiquitylation that function as regulators of CFTR stability and degradation have been discussed as promising therapeutic targets [70]. Mesenchymal stromal cells and induced pluripotent stem cells have been investigated in CF lung disease in vitro and in vivo, but have not been investigated in CFLD so far [71]. Finally, antifibrotic substances such as Farnesoid X Receptor (FXR) agonists, which target both the gut and the liver, and are centrally involved in bile acid homeostasis might be interesting candidates in the treatment of CFLD, but have not been investigated in this indication [72].
Funding: This research received no external funding.

Conflicts of Interest:
The authors declare no conflict of interest.