Management of liver disease and portal hypertension in cystic fibrosis: a review

ABSTRACT

Introduction

Cystic fibrosis (CF)-associated liver disease can significantly affect the quality of life and survival of people with CF. The hepatobiliary manifestations in CF are various, with focal/multilobular biliary cirrhosis more common in children and porto-sinusoidal vascular disease (PSVD) in young adults. Portal hypertensive complications, particularly bleeding from esophagogastric varices and hypersplenism are common, while liver failure is rarer and mainly linked to biliary disease.

Areas covered

This review explores current therapeutic options for CF-associated liver disease, presenting ongoing studies and new insights into parthenogenesis for potential future therapies.

Expert opinion

Monitoring for signs of portal hypertension is essential. Limited evidence supports ursodeoxycholic acid (UDCA) efficacy in halting CF liver disease progression. The effect of cystic fibrosis transmembrane conductance regulator (CFTR) modulators on liver outcomes lacks definitive data, since patients with CF-related liver disease were excluded from trials due to potential hepatotoxicity. A proposed approach involves using UDCA and modulators in early stages, along with anti-inflammatory agents, with further therapeutic strategies awaiting randomized trials. Prevention of portal hypertensive bleeding includes endoscopic sclerotherapy or ligation of esophageal varices. Nonselective beta-blockers may also prevent bleeding and could be cautiously implemented. Other non-etiological treatments require investigation.

KEYWORDS:

• Cystic fibrosis
• Cystic fibrosis-associated liver disease and portal hypertension
• CFTR modulators
• DHA
• multilobular biliary cirrhosis
• non-selective beta-blockers
• portosinusoidal vascular disease
• ursodeoxycholic acid

1. Introduction

Improved life expectancy and prolonged follow-up of people with cystic fibrosis (pwCF) have favored better characterization of the wide spectrum of conditions that may affect the hepatobiliary system. These include specific alterations ascribable to the underlying cystic fibrosis transmembrane conductance regulator (CFTR) defect, alterations of iatrogenic origin such as liver steatosis related to malnutrition and drug hepatotoxicity, the rare alterations reflecting the effects of a disease process that occurs outside the liver (such as hepatic congestion and common bile duct stenosis) and other disorders unrelated to CF but occurring in pwCF as well as in the general population [1].

CF-associated liver disease has emerged as a relevant comorbidity that may affect the quality of life and survival of affected patients. Indeed, liver disease continues to represent a significant cause of death, reported in 4,3% of CF patients deceased in 2022 in the US [2].

Despite the novel CFTR modulator therapies are potentially available for the majority of patients with responsive genotypes and have the potential of transforming the landscape of clinical care in CF [3], their use in pwCF with liver disease has been limited by their potential hepatotoxicity, thus impeding to evaluate whether they might prevent or delay its progression.

This review explores current therapeutic options for CF-associated liver disease, presenting ongoing studies and new insights into parthenogenesis for potential future therapies.

1.1. Literature search methodology

The main biomedical databases (i.e. Medline, Cochrane Library, and EMBASE) were searched to identify relevant studies published in English up to 31 December 2022 and periodically updated until March 2024. Reference lists of potentially eligible studies were also considered. The following terms were used for the search: cystic fibrosis, liver disease, treatment of liver disease, treatment of portal hypertension, liver transplantation.

1.2. The liver phenotype in CF

In this review article, based on the proposed phenotypic characterization of liver abnormalities in CF [4], the comprehensive term ‘liver involvement’ refers to the miscellaneous hepatobiliary manifestations of CF, carrying multifactorial, often not fully defined pathogenesis, and variable severity. Conversely, the term ‘liver disease’ is restricted to two different entities [5], both related to the underlying CFTR defect: focal and multilobular biliary cirrhosis and PSVD. The latter is a recently recognized disorder characterized by signs of portal hypertension (PHT) in the absence of cirrhosis, with obliterative portal venopathy as its pathognomonic histological lesion [5].

Focal biliary cirrhosis, until recently considered the only pathognomonic lesion related to CFTR dysfunction, is a biliary disease, with onset in childhood and potential progression to multilobular biliary cirrhosis and PHT. PSVD is a vascular disorder, characterized by the absence of cirrhosis, potential progression to presinusoidal PHT, and usually long-term preserved liver function. However, further studies are needed to gather information on prevalence, pathogenesis, and outcome of PSVD in pwCF. In both conditions, morbidity and mortality are mostly related to PHT. In contrast, liver failure is a rare and late event, mainly associated with the biliary disease.

Distinguishing biliary cirrhosis and PSVD eventually requires liver biopsy, which is invasive and has rarely been performed in pwCF, particularly in children. A diagnostic algorithm based on elastographic criteria has been recently proposed to distinguish the two disorders in pwCF with signs of PHT [6,7]. According to elastographic data, liver stiffness measurement (LSM) lower than 10 KPa or higher than 20 KPa rule out or rule in cirrhosis, respectively, thus suggesting PSVD in those with LSM lower than 10 KPa; spleen stiffness measurement (SSM) lower than 21 KPa exclude clinically significant PHT, and SSM > 50 kPa suggests PSVD. Finally, a spleen stiffness/liver stiffness ratio (SSM/LSM index) distinguishes the two disorders, being significantly higher in those with PSVD, with little overlap between the two forms [8].

These preliminary results seem to indicate PSVD as the prevalent liver disorder in pwCF, accounting for about 80% of the cases of those presenting with signs of PHT [6]. Importantly, this data is reinforced by reports showing that PSVD, not biliary cirrhosis, is the underlying disorder in 80–90% of CF patients undergoing liver transplantation [9–11]. Large studies are required to confirm such promising results.

This review will examine the therapeutic options presently available for liver disease and PHT in CF and report on those under study, based on the new insight regarding the pathogenesis that may pave the way to new potential targets for therapy.

2. Pathogenesis of liver disease in CF

In the hepatobiliary system, the CFTR protein is expressed at the apical membrane of the epithelial cells lining the biliary epithelium (cholangiocytes) where it regulates bile hydration and alkalization by maintaining chloride ion (Cl-) gradient that drives the secretion into the bile of bicarbonate by anion exchanger (AE2/SLC4A2) expressed either in the canaliculi or in the luminal membrane of bile duct epithelial cells [12].

Therefore, focal biliary cirrhosis, the typical hepatic lesion of CF has been classically considered a channelopathy, being the direct consequence of lack or dysfunction of CFTR protein in cholangiocytes, leading to inspissated biliary secretions, bile duct plugging, hepatocyte damage, inflammation, and progressive periportal fibrosis [13]. However, such a view is not fully supported by clinical and histologic data, since jaundice, bile inspissation, and cholestasis are rarely present in pwCF.

A more complex hypothesis involves the role of the gut–liver axis. Experiments on isolated cholangiocytes have shown that CFTR regulates epithelial innate immunity, specifically the toll-like receptor 4 (TLR-4)-dependent inflammatory responses, by inhibiting Rous sarcoma oncogene cellular homologue (Src) activity, while mutations in CFTR cause self-activation of Src, resulting in increased inflammatory cytokines and disruption of the epithelial barrier [14,15].

At the intestinal side, the lack of CFTR in the gut alters the normal composition of the microbiota, causing mucosal inflammation and increased intestinal permeability, thus ensuing translocation of bacteria or bacterial products into the portal circulation and excessive liver exposure to gut-derived endotoxins [16]. The synergistic effect of such changes and the altered cholangiocyte innate immunity would cause peribiliary inflammation and fibrosis [17]. This hypothesis seems to be confirmed by the study by Flass and colleagues [16], who found more macroscopic intestinal inflammatory lesions as well as slower bowel transit in pwCF with cirrhosis compared to age-matched pwCF without liver disease. In addition, differences in the fecal microbiome, although not significant, were observed in pwCF with cirrhosis, who had higher abundances of Firmicutes and lower abundances of Bacteroides that also correlated with macroscopic intestinal lesions.

A different pathogenic mechanism, probably related to CFTR expression in platelets and endothelial cells but not thoroughly yet investigated, would be involved in PSVD [18,19], with CFTR-related platelet activation and aggregation and endothelial damage triggering micro-thrombotic events [20,21].

3. Pharmacological management of liver disease in CF

The incomplete understanding of the pathogenesis of liver disease in CF along with the heterogeneous nature of the conditions that may affect the liver is probably the main reason explaining why treatments of proven efficacy to prevent liver disease and delay its progression are not yet available. Its complex pathogenesis that cannot be only attributed to lack or dysfunction of CFTR at the apical membrane of cholangiocytes, likely explains the limited efficacy of UDCA, the only drug implemented so far, which is exclusively focused on the biliary system.

Clinical trials addressing the efficacy of the recently developed CFTR modulators on CF-related liver disease are not yet available. However, according to the new insights, a pathophysiologic-directed therapeutic approach may be implemented. Such new pathogenic insights will likely prompt new targets for the management of CF-associated liver disease. Other treatment strategies have been proposed, but evidence for their efficacy is not yet available, since clinical trials with novel therapeutic agents have not been performed [Table 1].

Table 1. Pharmacological treatment for CF-associated liver disease.

Pharmacological treatment for CF-associated liver disease
Drug	Aim/Mechanism of action	Evidence of efficacy
CFTR modulators	Restore CFTR activity in different cell types (cholangiocytes, platelets, endothelial cells) level	Limited evidence in CF-associated liver disease
	Stimulate biliary secretion
UDCA/norUDCA	Reduce the bile acid pool hydrophobicity and hepatic inflammation	No evidence in advanced liver disease
DHA	Reduce hepatic inflammation	Experimental evidence in CF animal models
Probiotics	Manipulate the gut microbiota	No studies available in CF
Rifaximin	Modulates gut microbiome	No studies available in CF
	Reduces hepatic and systemic inflammation, bacterial translocation, and portal hypertension
NSBB Propranolol/Nadolol	Reduce portal pressure (decreased cardiac output, splanchnic vasoconstriction, reductionof portal and collateral blood flow	No evidence in CF-associated liver disease
Carvedilol	Reduces also intrahepatic resistance due to intrinsic vasodilatory activity
	More effective than propranolol in reducing portal pressure in patients with cirrhosis
Statins	Mechanisms involving fibrosis, inflammation, endothelial function, thrombosis and coagulation

DHA: docosahexaenoic acid; NSBB: Non-selective beta-blockers; UDCA: Ursodeoxycholic acid.

3.1. UDCA for CF-related liver disease

UDCA, a hydrophilic bile acid with choleretic properties, is at present the only available therapeutic approach. It has been prescribed for more than 30 years and is virtually devoid of serious side effects. The rationale for the use of UDCA in cholestatic liver diseases is to reduce bile viscosity, improve biliary secretion, and modify the bile acid pool composition by decreasing the proportion of toxic hydrophobic bile acids [22].

Concerns have been raised regarding the safety of long-term, UDCA treatment after Lindor et al. reported an increased risk of death and liver transplantation in patients with primary sclerosing cholangitis treated with high UDCA dose [23], possibly due to enhanced biotransformation of UDCA to the hepatotoxic secondary bile lithocholic acid. In a study involving 20 individuals with CF who were on long-term UDCA therapy, analysis of total and individual serum bile acids measured by stable-isotope dilution mass spectrometry, excluded the occurrence of excessive biotransformation to lithocholic acid which was present only in trace amounts [24].

However, its therapeutic role in CF-associated liver disease is controversial. Beneficial effects on liver biochemistry [25], hepatic excretory function and biliary drainage [26], liver histology [27], liver stiffness [28], fat absorption, and essential fatty acid status [29] have been documented, but only a few in the setting of randomized controlled studies. Moreover, there is limited information on the long-term effects of UDCA treatment on clinically relevant endpoints, such as liver transplantation and survival, mainly because of the short duration of the clinical trials (the maximum duration was 12 months) and the small sample size. In consideration of the long natural history of CF-associated liver disease, the reluctance of patients to stay on placebo as part of long-term outcome trials, and the relatively infrequent occurrence of clinically relevant events, it is unlikely that a prospective controlled trial will ever be performed.

A systematic Cochrane review highlighted the paucity of randomized controlled trials (RCTs) of UDCA use in CFLD and concluded that there is insufficient evidence to justify its routine use in CFLD [30].

Notably, a large longitudinal cohort study based on the UK registry (3417 cases) reported that UDCA had a positive effect on survival in patients with mild liver disease without cirrhosis [31]. Conversely, data from the French CF Modifier Gene Study (including 3328 patients), although largely based on retrospective observations, have recently suggested that UDCA treatment does not influence the development of severe liver disease with cirrhosis and PHT [32].

Similar conclusions were achieved by an international study involving 11 CF centers in Europe, Australia, New Zealand, and Russia and addressing whether UDCA prescription could prevent or delay the development of PHT and severe liver disease [33]. The study enrolled 1591 patients who followed-up for the occurrence of PHT with or without cirrhosis from diagnosis of CF up to 31 December 2016, 1192, from 9 prescribing centers and 399 from 2 non-prescribing centers. Patients followed-up in UDCA-prescribing centers did not have a lower incidence of severe liver disease and PHT as compared to those followed in centers not routinely prescribing UDCA, and the hazard ratio for the occurrence of PHT indicated no significant difference in relation to the UDCA practice of the center. Limitations of the study are inherent to its retrospective nature and to the fact that a more aggressive diagnostic approach and stricter follow-up adopted by centers which routinely prescribed UDCA may have led to increased recognition of severe liver disease.

Based on such controversial findings, the recent Cystic Fibrosis Foundation guidelines on liver disease recommend against UDCA treatment to prevent advanced liver disease [34].

Since there are currently no data for liver outcomes in pwCF who discontinue UDCA, indications for continuation or discontinuation of UDCA treatment for CF patients on UDCA should be considered with discretion by CF healthcare providers.

Due to limited efficacy of UDCA in various chronic cholestatic conditions, novel therapeutic approaches have been proposed including bile acid derivatives and nuclear and membrane receptor agonists. norUDCA is a side chain-shortened homologue of UDCA that does not undergo a full enterohepatic cycle but is passively absorbed from cholangiocytes, generating a HCO3-rich hypercholeresis [35]. Obeticholic acid, a semisynthetic farnesoid X receptor agonist adds choleretic, anti-fibrotic, and anti-inflammatory activity, and is presently approved as a second-line therapy for patients with Primary Sclerosing Cholangitis with incomplete response to UDCA [36]. However, there are safety concerns regarding this treatment for chronic liver disease, with pruritus reported in roughly 40% of the treated patients often leading to drug discontinuation. The efficacy and safety of both these compounds have not been tested in CF.

3.2. Treatment of liver disease with CFTR modulators: targeting the causative CFTR defect within the liver

A new class of drugs, named CFTR modulators have been developed over the last decade to correct the basic defect in CF and have been shown to partially restore CFTR protein function [37]. At present modulators are potentially prescribable to at least 70% of pwCF who live in high-income countries, providing significant benefits on lung disease and pulmonary exacerbations, nutritional status, quality of life and survival [38], whereas the effects on the gastrointestinal manifestations of the disease are smaller and less characterized [39].

Their clinical efficacy was demonstrated in phase 3 clinical trials, in patients with gating mutations (ivacaftor), homozygous for the F508del mutation (lumacaftor/ivacaftor and tezacaftor/ivacaftor), and in those compound heterozygous with at least one F508del mutation (elexacaftor/tezacaftor/ivacaftor) [37]. There are at present four CFTR modulator therapies with market authorization for the treatment of pwCF. The triple combination elexacaftor/tezacaftor/ivacaftor (ETI) induced marked clinical improvement in pwCF with at least one F508del mutation and is referred to as highly effective modulator therapy potentially available for a large population of pwCF [37].

However, registration studies of CFTR modulators focused almost exclusively on pulmonary and nutritional outcomes and the presence of clinically relevant liver disease was a consistent exclusion criterium for their potential hepatotoxicity. Therefore, their efficacy and safety on CF-associated liver disease needs evaluation in real-world studies.

In this context, a few studies have recently addressed the issue of efficacy of CTFR modulators in pwCF and liver disease, with controversial results. A study by Drummond et al. [40] including 28 adolescents with CF and abnormal liver biochemistry before starting lumacaftor–ivacaftor, showed that liver function tests might improve during treatment. Levitte et al. conducted a retrospective, single-center analysis of 84 children and adolescents with CF-associated liver involvement treated with lumacaftor/ivacaftor and/or ETI therapy, focusing on alterations in liver function tests and biomarkers of liver fibrosis [41]. Significant decreases in liver function tests, APRI (aspartate aminotransferase-to-platelet index) and GPR (gamma-glutamyl transferase-to-platelet ratio) were observed during lumacaftor/ivacaftor treatment, but not in patients treated with ETI.

With regard to liver stiffness measurement after triple combination (ETI), Calvo et al., reported a significant reduction in liver stiffness in all pwCF with liver disease during the first month of treatment, which may indicate a rapid improvement in biliary secretion with a lower viscosity that might lead to amelioration of bile duct damage [42]. Conversely, a study by Schnell et al. [43] documented an increasing liver stiffness and altered bile acid metabolism in 20 children and young adults shortly after initiation of ETI treatment, as a possible sign of liver damage. This was not observed in a recent study in adult pwCF, showing a reduction in liver stiffness in 74 patients with the initial highest results and liver nodularity [44].

Thus, highly effective modulator therapies do not seem to consistently improve markers of hepatic fibrosis, at least in the short-term use (<1 year). Additional studies on a larger number of pwCF with liver disease using prospectively state-of-the-art techniques to assess liver disease are necessary to establish the effectiveness of CFTR modulators for both the treatment and prevention of liver disease in CF.

On the other hand, CFTR modulators can induce transient liver function test abnormalities in 5–15% of the patients, though most patients do not require treatment cessation, transient interruption, and succeeding re-challenge being usually adequate [43]. Interestingly, prospective serial monitoring of liver tests at a large single adult CF center documented a mild, likely insignificant increase of standard liver function tests after 3 months that did not increase further in the vast majority of patients [45]. Consistent liver function test trends were also noted in our pilot study, which involved 56 patients. Elevated ALT values exceeding three times the upper limit of normal occurred in three patients, while GGT values surpassed the limit in one patient, and bilirubin levels were more than double the upper limit in four patients, all but one with concomitant Gilbert syndrome. Although no patients discontinued treatment, three reduced ETI dose. Most elevations in liver function tests occurred within the initial month of treatment and persisted above the upper limit thereafter [46].

Of note, isolated elevation of serum total bilirubin after initiation of ETI are often associated with Gilbert’s syndrome and genetic testing for this relatively common syndrome should be considered in these patients, since appropriate diagnosis may avoid unnecessary interruption in this therapy with significant health benefits in CF [47].

However, much more rarely clinically relevant drug-induced liver injury with jaundice and even irreversible decompensation may occur several months after the start of ETI treatment in pwCF without previously diagnosed liver disease [48,49].

Therefore, in clinical practice, it is recommended that pwCF should have liver function tests monitored at least every 3 months for the first year of treatment; thresholds for dosing interruption have been recently outlined along with specific recommended actions for monitoring [50].

With regard to pwCF with liver disease, in a French multicenter study by Burgel et al. [51], pwCF with compensated cirrhosis had a relatively good tolerance to lumacaftor–ivacaftor treatment and the presence of cirrhosis/PHT was not associated with an increased risk of treatment discontinuation. In fact, only 1 out of 42 pwCF with cirrhosis/PHT that were included in this study had liver enzyme elevation resulting in study discontinuation. The authors concluded that Lumacaftor-Ivacaftor may be prescribed under close monitoring in patients with compensated cirrhosis, because the benefits of CF lung disease likely appear to outweigh the liver-related risk.

Assessment of the safety and pharmacokinetics of CFTR modulators is necessary to better guide dosing recommendations in pwCF with impaired hepatic function. A phase 1 open-label study was conducted to evaluate the safety and pharmacokinetics of the three-drug, fixed-dose combination of ETI and their metabolites, in 11 subjects without CF with moderate hepatic impairment (cirrhotic patients, Child-Pugh class B) and in 11 healthy controls [52]. Despite all subjects received half the standard daily dose of ETI for adult pwCF with normal hepatic function, on day 10 the exposures (reflected by mean values of the area under the curve during the dosing interval) of elexacaftor, its major metabolite M23-elexacaftor, tezacaftor, and ivacaftor were higher in subjects with moderate hepatic impairment compared with matched healthy subjects. It was concluded that in pwCF with moderate hepatic impairment, ETI should be prescribed at a reduced dose [53], it should not be used in subjects with severe hepatic impairment, whereas no dose adjustment is needed in patients with mild hepatic impairment [52].

Another small series was described of seven pwCF with cirrhosis Child-Pugh A or B in whom a strategy of stepwise dose increase of ETI (in 4 pre-defined dosing steps), with strict clinical, biochemical and therapeutic drug monitoring allowed doses to be administered that did not induce clinical adverse effects or increases in serum liver enzymes and produced exposure comparable with pwCF without hepatic impairment receiving the regular dose [54]. Despite the limitations of this small case series, stepwise elevation of ETI dose may allow tolerable introduction of this therapy in pwCF and cirrhosis Child-Pugh A and possibly B.

Finally, a study reported the experience on 15 adult pwCF prescribed ETI who underwent dose reduction due to adverse events during ETI therapy (mainly consisting in ALT increase) with the aim of maintaining therapeutic efficacy while resolving the adverse events. Full physiologically based pharmacokinetic models of ETI were developed, incorporating physiological information and drug-dependent parameters, and were then used to predict lung concentrations of ETI at steady-state. Clinical stability without significant changes in ppFEV1 after dose reduction was observed in all patients and resolution or improvement of AEs occurred in 13 of the 15 cases, suggesting that this strategy may be effectively used for maintaining therapeutic efficacy at reduced dose [55].

Overall, given the above-described controversial results and caveats, the recent Cystic Fibrosis Foundation guidelines could not recommend for or against the use of CFTR in pwCF with compensated cirrhosis and recommended against CFTR modulator treatment in patients with decompensated liver disease, defined by INR > 1.5, abnormal direct bilirubin, low albumin, refractory ascites, or encephalopathy [34].

However, the issue of highly effective modulators use in patients with advanced liver disease requires further studies on an adequate number of pwCF with prolonged follow-up evaluation. For the time being, treatment may be prescribed to patients with compensated cirrhosis at a reduced dose and not recommended in those with decompensated disease.

ETI has so far rarely been prescribed in the CF lung transplant population and further studies are needed to determine the impact on long-term post-transplant outcomes given the potential for drug interactions [55].

Finally, it should be considered that even if the presently available CFTR modulators will prove to be effective in the treatment and prevention of liver disease in pwCF, a recent international study has documented that a significant proportion of pwCF with advanced liver disease would be ineligible for ETI or ivacaftor therapies, and could not take advantage of the potential benefits of these new treatments. The consequences for the ineligibility of patients with extreme liver phenotype may be even more clinically relevant because of their poorer disease risk profile [56].

3.3. Omega-3 fatty acids and docosahexaenoic acid (DHA)

Docosahexaenoic acid (DHA) is an omega-3 polyunsaturated fatty acid derived from the essential fatty acid alpha linoleic acid. DHA plays multiple roles in human development, particularly for brain structure and function, and also later in life due to its anti-inflammatory properties.

The presence of essential fatty acid deficiency in CF was reported since many years both in pwCF and in animal models. Freedman et al. [57] described a CFTR-/- mouse model of CF (due to exon 10 knockout) with important abnormalities of fatty acid metabolism, consisting of a membrane lipid imbalance with an increase in phospholipid-bound arachidonic acid and a decrease in phospholipid-bound DHA. This lipid imbalance was observed in organs pathologically affected by CF including lung, pancreas, and ileum, and was not secondary to impaired intestinal absorption or hepatic biosynthesis of DHA. In addition, an association between DHA deficiency and abnormal intestinal and pancreatic morphology was found [57] and the observed biochemical and morphological abnormalities could be reversed by pharmacological doses of DHA, suggesting that certain phenotypic manifestations of CF may result from remediable alterations in phospholipid-bound arachidonic acid and DHA levels. The same group of investigators later reported alterations in fatty acids similar to those found in CF-knockout mice in nasal and rectal-biopsy specimens from patients with CF [58].

However, in another congenic CFTR -/- murine model that develops CF-like pathology in all organs, DHA did not demonstrate any significant effect on the lung, pancreas, or ileum but was associated with remarkable histological liver changes, with reduction of periportal inflammation, without changes of the obstructive morphologic abnormalities [59].

Thus, the primary effects of DHA were to reduce inflammation rather than influence CFTR function. Inhibition of cytokines and/or eicosanoid metabolism and release of endogenous inhibitors of inflammation by DHA may account for the observed anti-inflammatory effect on the liver.

Subsequent studies have also explored the protective effect of DHA from liver damage. The liver content of DHA was evaluated in patients with cirrhosis (not CF-related) and in controls by liquid chromatography-tandem mass spectrometry [60]. Cirrhotic livers showed a marked depletion of DHA, which correlated with the progression of liver disease. In the same study, the authors utilized a murine model of acute liver injury to evaluate the effect of DHA on the response of hepatic stellate cells (the main producers of collagen in the liver) to pro-fibrogenic stimuli and demonstrated a strong anti-fibrogenic effect achieved through the attenuation of activation and fibrogenic response of these cells. In addition, DHA appeared to inhibit fibrogenesis more intensely than other omega-3 fatty acids.

Of interest, however, is the observation by Van Biervliet S et al. [61] carried out to evaluate the relation of clinical parameters and genotype with the serum phospholipid fatty acid composition in 104 CF patients with a median age of 15 years. CF patients, particularly those with a severe genotype, had significantly lower DHA and linoleic acid and higher dihomo γ-linolenic acid, oleic acid, and mead acid. There was no relation between serum fatty acid composition with nutritional status, caloric intake, pancreatic function, gender, pulmonary function, Pseudomonas colonization or diabetes mellitus, but in CF patients with liver disease, DHA was significantly lower than in patients of the same genotype with a normal liver.

More recently, similar differences in fatty acid composition of serum phospholipids between CF patients with and without cirrhosis were confirmed [62]: the serum levels of n-3 fatty acids including DHA were lower in cirrhotic compared to non-cirrhotic patients and these differences were not explained by dietary fat intake. The potential therapeutic benefits of DHA in CF-associated liver disease remain to be explored [63].

Finally, EFA deficiency itself can also cause liver abnormalities. In a prospective study on the natural history of liver disease in CF, EFA deficiency occurred more often in patients with marked steatosis on liver biopsies [64], the degree of steatosis being inversely related to molar percentage of LA in serum phospholipids at the time of liver biopsy. Therefore, correction of EFA status may reduce steatosis and also prevent possible progression to more severe hepatic lesions.

Overall, these studies suggest that dietary DHA supplementation and adequate rebalance of essential fatty acid deficiency could be beneficial in patients with CF-associated liver disease [63], thus underlying the need to explore such possibility in long-term, multicenter, possibly randomized studies.

4. Management of portal hypertension in CF

The paucity of studies useful for evidence-based recommendations [65] explains why specific guidelines addressing the prevention of portal hypertensive gastrointestinal bleeding are not available yet. Therefore, the suggestion is to keep to the guidelines for the management of PHT in the non-CF population, if possible.

4.1. Pharmacological management of portal hypertension

4.1.1. Non-selective beta-blockers

The pharmacological management of primary and secondary prophylaxis of gastrointestinal bleeding in CF patients might include NSBB, like propranolol or nadolol, at the maximum tolerated dose. Such treatment lowers cardiac output and causes vasoconstriction in the gastrointestinal tract, thereby reducing portal and collateral blood flow. NSBB prevents bleeding from esophageal varices in more than 50% of the non-CF patients with cirrhosis and medium or large esophageal varices [66].

However, there are concerns regarding NSBB treatment in pwCF because of their potential adverse effects on pulmonary disease. At present, the efficacy and safety of NSBB in preventing variceal bleeding has not been evaluated in pwCF. Therefore, no recommendation pro or cons NSBB treatment in pwCF with PHT and gastroesophageal varices at bleeding risk can be conveyed.

In recent years, Carvedilol, a NSBB with added intrinsic vasodilatory activity that also reduces intrahepatic resistance through alpha 1 blockade, has been shown to be more effective than propranolol in reducing portal pressure in patients with cirrhosis [67–69].

Therefore, carvedilol is, at present, the first-choice drug for the pharmacological management of PHT, at a daily dose of 12.5 mg (6.25 mg twice a day). Although not yet evaluated in pwCF, carvedilol, if tolerated, might be an effective option for preventing variceal bleeding in such population and worthy of appropriate clinical evaluation. Higher doses could cause more side effects, in particular arterial hypotension, without further therapeutic benefit, as shown in patients with advanced chronic liver disease unrelated to CF. Whether lower doses might be sufficiently effective in CF patients, while minimizing side effects, deserves evaluation in clinical studies.

4.2. Non-etiological treatments for portal hypertension

Besides PHT, bacterial translocation, and hepatic and systemic inflammation drive the risk of decompensation in patients with advanced liver disease. In recent years, studies addressing these targets have been performed, mainly aimed at increasing the effects of NSBB [70,71]. The novel non-etiological treatments mature for implementation in clinical practice are statins, albumin, aspirin, anticoagulants, and rifaximin. Though no such treatment has been attempted in pwCF, it is likely that, following their next widespread utilization in chronic liver disorders, some of them could also be implemented in CF-related liver disease.

4.2.1. Statins

Statins could be beneficial in chronic liver diseases, well beside their cholesterol-lowering effect, through pleiotropic mechanisms involving inflammation, fibrosis, endothelial function, thrombosis, and coagulation [72–75]. Concerns on their potential hepatotoxicity have so far precluded their clinical implementation. However, in most chronic liver disorders, the benefits likely outweigh the risks, and clinical studies are needed.

In pwCF with liver disease, more often non-tolerant to NSBB or Carvedilol, statins could be a therapeutic option, either alone or in addition to low dose NSBB. However, in experimental models of cirrhosis (bile duct ligated rats) or non-cirrhotic portal hypertension (partial portal vein ligated rats), atorvastatin showed positive effects on portal pressure, shunt flow and angiogenesis only in cirrhotic rats, and contrary effects in non-cirrhotic portal hypertensive rats. This argues in favor of their possible use in CF patients with biliary cirrhosis, but against their use in pwCF and PSVD [76].

4.2.2. Albumin

The biological functions of albumin include maintaining colloid osmotic pressure, antioxidative properties, endothelial stabilization, and immunomodulatory functions such as reduction of cytokine-induced tissue injury, inhibition of pro-inflammatory cytokines by bacterial DNA-stimulated immune cells, and anti-inflammatory action without impairing immune response [77,78]. Given the relevant role of systemic inflammation, oxidative stress, circulatory and immune dysfunction in the development of complications of cirrhosis, guidelines recommend albumin infusion for the treatment of selected complications of cirrhosis, such as hepatorenal syndrome, spontaneous bacterial peritonitis, and large volume paracentesis [79–81]. A meta-analysis of randomized clinical trials suggests that albumin decreases mortality in patients with cirrhosis [82]. Whether pwCF with advanced liver disease might benefit from albumin infusion to treat or prevent spontaneous bacterial peritonitis or, possibly, other infections, deserve studies.

4.2.3. Aspirin

Beyond their hemostatic properties, platelets have a role in hepatic injury and fibrosis, share pro-inflammatory properties with liver sinusoidal endothelial cells and Kupffer cells, and drive a pro-hepatocarcinogenic environment. Therefore, antiplatelet drugs may have a role in the treatment of PHT [83–86].

4.2.4. Anticoagulants

The historical paradigm of liver disease as a hypocoagulable condition with increased bleeding risk has been challenged in the last two decades, and the current knowledge considers hemostasis in cirrhosis as a rebalanced, though fragile condition, resulting from the interaction of both decreased procoagulants (with the notable exception of Factor VIII and von Willebrand factors that are increased) and anticoagulants of hepatic synthesis. The resulting fragile equilibrium may be tipped toward either bleeding or thrombosis by accidental events such as infection [87].

However, in advanced liver disease, the equilibrium is shifted toward a prothrombotic state, as shown by the frequent occurrence of portal vein thrombosis [87]. Growing evidence suggests that the hypercoagulable state of cirrhosis enhances fibrogenesis while hypocoagulability decreases it [88–90], and that anticoagulation prevents portal vein thrombosis and liver decompensation [91] without increasing the risk of bleeding. Direct oral anticoagulants are also promising treatments for PHT, with the advantage of oral administration. A multicenter prospective randomized trial aimed at evaluating the effect of Rivaroxaban on survival and development of complication of PTH is in progress (CIRROXABAN, clinical trials identifier NCT02643212 [92]). No data is presently available. However, whether low-dose low molecular weight heparins or direct oral anticoagulants will convincingly prove safe and effective in people with cirrhosis and PTH, the evaluation of their effectiveness and safety in pwCF and advanced liver disease with PTH might be the next step. However, the issue of safety in pwCF, based on the risk of hemoptysis, should be considered.

4.2.5. Rifaximin

Rifaximin, an oral non-systemic antibiotic with minimal gastrointestinal absorption and broad-spectrum antibacterial coverage is widely used for preventing or treating hepatic encephalopathy [93]. Through modulation of the gut microbiome, rifaximin decreases hepatic and systemic inflammation, bacterial translocation, and PHT [94–97]. Benefits on complications other than hepatic encephalopathy, though not conclusive, due to retrospective studies on small RCT, suggest protection from infections and further decompensation [94–97]. Further, large studies are in cirrhosis patients and specific studies in pwCF and advanced liver disease are needed.

4.3. Non-pharmacological management of portal hypertension

Since NSBB may likely not be tolerated in pwCF because of the pulmonary disease, several non-pharmacological treatments are valuable options to prevent bleeding in those with high-risk varices (i.e. large varices or even small varices with red wall signs or gastric varices), or to prevent recurrent bleeding [98]. Such options include endoscopic treatments, as esophageal variceal ligation (EVL) or sclerotherapy, and radiological or surgical procedures aimed at reducing PHT and/or counteracting hypersplenism, as briefly outlined below.

4.3.1. Endoscopic treatments

EVL is superior to sclerotherapy because it is more effective, it does not require repeated antibiotic prophylaxis, and anesthesia is required only in children and in patients with acute variceal bleeding [99,100]. EVL must be performed until eradication of esophageal varices, usually requiring two to three sessions, three to four weeks apart to be achieved. Sclerotherapy, although less effective and burdened by more frequent adverse events, is an alternative option in small children, due to the lack of banding devices adequately sized for pediatric endoscopes.

4.3.2. Surgical or radiological treatments

Possible surgical options are portosystemic shunts, which are currently performed rarely, as transjugular intrahepatic portosystemic shunt (TIPS) is a less invasive derivative procedure, and the polytetrafluoroethylene (PTFE)-covered stents now widely adopted to reduce the incidence of TIPS dysfunction.

TIPS is an effective treatment for the complications of PHT and may delay or avoid the need for transplantation, in CF patients with advanced liver disease. Especially in the context of a pediatric population, the potential impact on neurological development should be carefully considered when contemplating TIPS procedures. In fact, patients undergoing TIPS are notably prone to developing hepatic encephalopathy, with the incidence ranging from 10% to 50% [101].

Porto-systemic shunting, either TIPS or surgical, might be preferable over liver transplantation in patients with CF-related noncirrhotic portal hypertension, given the absence of cirrhosis and the usually long-term preserved liver function [65,102,103].

Other surgical options, worthy of consideration particularly in patients with severe hypersplenism, are total or partial splenectomy, alone or associated with splenorenal shunt [104,105].

4.3.3. Orthotopic liver transplantation

Finally, orthotopic liver transplantation (OLT) is an established option in pwCF with end-stage liver disease, because it significantly increases survival [106], though selection criteria and optimal timing for OLT are a matter of debate. In the setting of evaluation for OLT, it is crucial to assess whether OLT alone is required or a multiorgan transplantation is more appropriate, carefully evaluating the severity of pulmonary and pancreatic involvement. Indeed, the severity of cardiorespiratory disease is variable, and patients may require multiple-organ transplantation or be too ill for any procedure. As expected, the outcomes of combined liver-lung or liver-heart-lung transplantation are worse than those of OLT alone [107,108].

Of note, poor growth and nutritional status are associated with high post-transplant mortality rate [106]. Moreover, studies have not shown consistent nutritional improvement following OLT, thus not qualifying poor growth or nutritional status ‘per se’ as OLT indications [106]. Lung function may either improves or worsen after OLT, and any lung function improvements, if occurring, appear to vanish within three years of OLT [109].

Long-term outcomes of transplantation are acceptable, though lower than that of individuals undergoing OLT for other etiologies. However, a survival benefit was documented in transplanted patients versus those who remained on the waiting list [108].

5. Conclusions

Over the last few years, it has become clear that the term liver disease in CF includes two entities both related to the CF basic defect, a biliary disease developing in childhood that may progress to multilobular cirrhosis and the more recently described, PSVD, leading to non-cirrhotic portal hypertension, most often in adulthood. PHT is the most common complication in both conditions, but the optimal therapeutic approach may differ.

With regard to pharmacological treatment of PHT, NSBB (Carvedilol better than propranolol or nadolol), if tolerated, may prevent portal hypertensive bleeding, but there are no studies on this relevant issue and these agents have been under-used in CF, favoring non-pharmacological treatments that are certainly more invasive and not devoid of side effects. These include endoscopic treatment of esophageal varices and surgical procedures (portosystemic shunts, presently seldom performed, and total or partial splenectomy), TIPS and liver transplantation in cases of liver failure.

Due to the limitations of long-term studies so far carried out, the overall efficacy of UDCA in liver disease in CF remains unclear, consequently, indications for continuation or discontinuation of treatment for CF patients with cirrhosis and PH presently on UDCA should be considered with discretion by CF healthcare providers.

The use of CFTR modulators also lacks definitive data regarding its impact on pwCF with liver disease, and their potential benefits need to be further investigated in large, prospective studies. If liver-related benefits are confirmed, it is crucial to remember that a considerable portion of pwCF and liver disease might be excluded from starting modulator treatment due to genotype ineligibility, drug costs, and emergence of side effects [37]. In this respect, research is ongoing to develop additional therapies to correct the underlying genetic or molecular defect in order to address the issue of equity in health care for all people with CF.

In this review article, we describe the possible therapeutic options for CF-associated liver disease, based on most recent studies, and highlight the lack of evidence-based pharmacological treatments to improve/delay its progression. No major advances occurred despite liver disease is a major comorbidity in pwCF. The paucity of methodologically adequate studies likely depends on the long natural history of CF-related liver disease, and the relative rarity of liver-related events. Therefore, adequately powered multicenter studies addressing the effects of the more promising therapies currently under evaluation in other advanced chronic liver disorders are urgently needed.

6. Expert opinion

Two phenotypes of liver disease occur in pwCF, with different pathophysiology and natural history, biliary cirrhosis being the main phenotype in childhood, and PSVD being more prevalent in youth and adults. Whether such disorders require different treatments to prevent their progression and the development of complications deserves future studies.

Patients with PSVD may be at higher risk of complications from PHT, mainly bleeding from esophageal or gastric varices, liver failure occurring rarely. Since PSVD appears to be the most prevalent condition of pwCF undergoing OLT, and the suboptimal OLT outcome given the pulmonary disease, less radical procedures aimed at decreasing PHT, such as TIPS or derivative surgery could be more adequate in such patients with preserved liver function. Hence, distinguishing biliary cirrhosis and PSVD is relevant, and potentially involves the reassessment of OLT indication. Efforts should be pursued to promote large studies also using noninvasive diagnostic tools such as liver and spleen stiffness measurement to achieve this goal [6].

CFTR modulators prevent lung function decline and lung failure making lung transplantation unlikely. However, it’s important to note that CFTR modulators might be contraindicated in the presence of liver disease (e.g. advanced liver disease). Consequently, there arises a speculation that certain individuals may necessitate OLT to facilitate the safe administration of CFTR modulators, thereby optimizing both their therapeutic benefits and quality of life.

At present, there are no medical treatments of proven efficacy to improve and delay progression of liver disease in CF. The only therapeutic option so far available is the oral administration of UDCA, which has been widely used in CF patients and shown to exert beneficial effects on different aspects of liver disease, including liver biochemistry, histopathological alterations, early hepatic ultrasonographic changes and to reduce liver stiffness in CF patients with mild liver disease.

No significant side effects related to the long-term use of this drug have ever been reported, and extensive analysis of serum bile acid composition in CF patients on long-term UDCA treatment failed to show significantly higher concentrations in the hepatotoxic lithocholic acid.

Unfortunately, due to the long natural history of CF-associated liver disease, there are limited data on the effectiveness of UDCA on long-term outcomes including death or need for liver transplantation, but large observational studies did not provide evidence of the efficacy of UDCA in halting the progression of liver disease to cirrhosis and portal hypertension. On the other hand there are no data on discontinuation of UDCA treatment, although in clinical practice we have repeatedly observed biochemical alteration shortly after its interruption. Asymptomatic patients with early-stage liver disease are more likely to benefit from UDCA administration, but there are no long term data. Regular and systematic screening for liver involvement enables early introduction of UDCA therapy in affected cystic fibrosis patients, reduces the development of severe liver disease and leads to a significant and persistent improvement in serum liver tests [110]. A prophylactic study evaluating the efficacy of UDCA in at risk CF patients is required to establish its role in the prevention of liver disease.

Whether CFTR modulators are safe and effective in pwCF with liver disease is unknown, since such patients were excluded from clinical trials. In pwCF with abnormal liver tests before starting modulators, liver biochemistry may improve, but data on fibrosis biomarkers are limited and controversial. A few preliminary studies have shown that therapeutic drug monitoring may represent a useful strategy to manage the hepatotoxic effects of CFTR modulators and allowed doses to be adjusted to the real need of the individual patient, without precluding the positive effects of modulators on lung disease and nutritional status. This approach should be implemented, when analytical assays will become widely available.

Conversely, indices of cytolysis and/or cholestasis may worsen after starting CFTR modulators, though such changes might not be clinically relevant in many cases. At present, guidelines argue against treating with modulators in pwCF with advanced liver disease because of the risk of hepatotoxicity. Future studies addressing this issue will better clarify indications and contraindications of modulators in such patients. Regarding pwCF submitted to OLT, the recent Cystic Fibrosis Foundation Guidelines support the use of modulators, but point out the need for close monitoring to assess potential drug interactions and adverse effects [34].

As for PHT management, carvedilol, with its added vasodilatory activity, might be more effective than other NSBB, like propranolol on nadolol, which only targets splanchnic hyperemia.

Novel non-etiological treatments aimed at reducing the intrahepatic resistance, addressing endothelial dysfunction, liver fibrosis, liver and systemic inflammation, and bacterial translocation, are under evaluation in patients with chronic liver disorders. Statins, albumin, anticoagulants, aspirin, rifaximin, and fecal microbiota transplant are currently mature to be included in the therapeutic armamentarium for chronic advanced liver disorders. None of these has been evaluated in pwCF, but their therapeutic potential calls for a future evaluation in clinical studies and, in case of favorable results, for a subsequent implementation in clinical practice.

In conclusion, an accurate diagnosis of the underlying CF-related disease might be possible, given the advent of new diagnostic devices and algorithms. A possible synergistic treatment approach involves UDCA in the early stage of liver disease, modulators, anti-inflammatory agents, and CFTR protein stabilizers, with additional therapeutic strategies awaiting randomized controlled trials. Finally, optimizing nutritional status is of key importance for pwCF either with cirrhosis or PSVD, which are to be considered highly catabolic conditions [111].

Article highlights

• Cystic Fibrosis (CF)-related liver disease encompasses two different entities, with the CFTR protein’s involvement in bile regulation and emerging insights suggesting a gut–liver axis role in the pathogenesis of the biliary disease; another pathogenic mechanism possibly tied to CFTR expression in platelets and endothelial cells is involved in the vascular disease (portosinusoidal vascular disease) with development of non-cirrhotic portal hypertension.

• While the primary current used drug ursodeoxycholic acid (UDCA) faces controversy and lacks strong evidence on long-term efficacy, new insights suggest a potential for pathophysiologic-directed therapeutic approaches, such as a diagnostic algorithm utilizing elastographic tools, with preliminary data indicating portosinusoidal vascular disease (PSVD) as the predominant liver disorder in pwCF with portal hypertension, highlighting a need for further investigations and potential new targets in the management of CF-associated liver disease.

• CFTR modulators raise concerns about liver impact, with conflicting findings on hepatic fibrosis markers, occasional and transient liver function abnormalities, and rare instances of serious drug-induced liver injury; recent guidelines recommend against its use, in pwCF with advanced liver disease and treatment with a reduced dose in those with liver disease of moderate severity.

• Docosahexaenoic acid (DHA) and essential fatty acid (EFA) status play complex roles in CF-associated liver disease, with DHA potentially alleviating inflammation and fibrosis. Still, further exploration through long-term, multicenter, and randomized studies is necessary.

• Evidence scarcity hampers specific guidelines for preventing portal hypertensive gastrointestinal bleeding in CF, prompting adherence to non-CF portal hypertension management guidelines. Pharmacologically, nonselective beta-blockers, carvedilol, and statins could be considered for pwCF, while non-pharmacological options, such as endoscopic and surgical interventions, are valuable for preventing or managing bleeding in CF patients.

• Orthotopic liver transplantation remains a feasible option in end-stage liver disease, but the optimal approach requires further research and careful assessment of associated pulmonary and pancreatic conditions. Treatment of organ transplant recipients with CFTR modulators under strict monitoring is encouraged.

Abbreviations

APRI aspartate aminotransferase-to-platelet index; CF Cystic fibrosis; CFTR Cystic fibrosis transmembrane conductance regulator; DHA Docosahexaenoic acid; EFA Essential fatty acid; ETI Elexacaftor-tezacaftor-ivacaftor; EVL Esophageal variceal ligation; GPR gamma-glutamyl transferase-to-platelet ratio; INR International Normalized Ratio; LA linoleic acid; LSM liver stiffness measurement; NSBB Non-selective beta-blockers; OLT Orthotopic liver transplantation; PHT Portal hypertension; PTFE polytetrafluoroethylene; PSVD Portosinusoidal vascular disease; pwCF People with cystic fibrosis; SSM spleen stiffness measurement; UDCA Ursodeoxycholic acid; TIPS Transjugular intrahepatic portosystemic shunt; US Ultrasound.

Declaration of interest

The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or material discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or mending, or royalties.

Reviewer disclosures

Peer reviewers on this manuscript have no relevant financial or other relationships to disclose.

Additional information

Funding

This paper was not funded.