Progress of personalized medicine of cystic fibrosis in the times of efficient CFTR modulators

Abstract

Background

Cystic fibrosis (CF) is a systemic disorder of exocrine glands that is caused by mutations in the CFTR gene.

Main body

The basic defect in people with CF (pwCF) leads to impaired epithelial transport of chloride and bicarbonate that can be assessed by CFTR biomarkers, i.e. the β-adrenergic sweat rate and sweat chloride concentration (SCC), chloride conductance of the nasal respiratory epithelium (NPD), urine secretion of bicarbonate, intestinal current measurements (ICM) of chloride secretory responses in rectal biopsies and in bioassays of chloride transport in organoids or cell cultures. CFTR modulators are a novel class of drugs that improve defective posttranslational processing, trafficking and function of mutant CFTR. By April 2025, triple combination therapy with the CFTR potentiator ivacaftor (IVA) and the CFTR correctors elexacaftor (ELX) and tezacaftor (TEZ) has been approved in Europe for the treatment of all pwCF who do not carry two minimal function CFTR mutations. Previous phase 3 and post-approval phase 4 studies in pwCF who harbour one or two alleles of the major mutation F508del consistently reported significant improvements of lung function and anthropometry upon initiation of ELX/TEZ/IVA compared to baseline. Normalization of SCC, NPD and ICM correlated with clinical outcomes on the population level, but the restoration of CFTR function was diverse and not predictive for clinical outcome in the individual patient. Theratyping of non-F508del CF genotypes in patient-derived organoids and cell cultures revealed for most cases clinically meaningful increases of CFTR activity upon exposure to ELX/TEZ/IVA. Likewise, every second CF patient with non-F508del genotypes improved in SCC and clinical outcome upon exposure to ELX/TEZ/IVA indicating that triple CFTR modulator therapy is potentially beneficial for all pwCF who do not carry two minimal function CFTR mutations. This group who is not eligible for CFTR modulators may opt for gene addition therapy in the future, as the first-in-human trial with a recombinant lentiviral vector is underway.

Future directions

The upcoming generation of pwCF will probably experience a rather normal life in childhood and adolescence. To classify the upcoming personal signatures of CF disease in the times of efficient modulators, we need more sensitive CFTR biomarkers that address the long-term course of airway and gut microbiome, host defense, epithelial homeostasis and multiorgan metabolism.

Background

Prologue: The Osler – Garrod contrast

Barton Childs (1916–2010) was an American pediatrician and geneticist, who defined the field of genetic medicine and provided the best rationale for its existence. In his 1999 book, Genetic Medicine: A Logic of Disease [1], he introduced his concept of disease as disturbed homeostasis due to interacting genetic and societal factors. Taking the contrast between William Osler (1849–1919) and Archibald Garrod (1857–1936), two of the most influential physicians in the early twentieth century, Childs argued that in the future, all medicine, or medical theory, must be based on the individuality of gene – environment interaction. Osler, the activist, conjured with facts; Garrod, a contemplative man, with ideas. The Oslerian spirit of medicine shaping our todays consensus guidelines (’Leitlinien’) reflects the emphasis on the disease and how its effects are to be reversed. The patient, who is perceived as representative of the class of people with the disease at hand, might be anybody. In other words, Oslerian thinking is organized about treatment and management. In contrast, Garrod saw the patient as a less well adapted product of evolution and the disease as a consequence of a unique individual’s encounter with an environment for which he was uniquely unfit. Susceptibility to disease is a consequence of the chemical and biological individuality among patients. Thus, Garrodian thinking paved the way to our present concept of personalized medicine. Childs concluded: ‘No one would deny that Osler was the hero of the medicine of the twentieth century. It is likely that Garrod will be the icon of the twenty-first’ [1].

Here, writing about people with cystic fibrosis (pwCF) will be inspired by Garrodian thinking decorated by Oslerian notes if necessary.

In brief: Cystic fibrosis – status 2025

Cystic fibrosis (CF) is a systemic disorder of exocrine glands that primarily affects the respiratory, gastrointestinal, hepatobiliary and reproductive tracts [2, 3]. CF is an autosomal recessive disease affecting more than 190,000 people worldwide [4] that is caused by two mutations in trans in the Cystic Fibrosis Transmembrane Conductance Regulator (CFTR) gene [5]. The universal feature of CF organ damage caused by CFTR deficiency is the buildup of mucus that is thick, tenacious, and dehydrated [6]. Thanks to continuously improved symptomatic treatment during the last five decades [7–9] this lethal pediatric disease has been transformed into a chronic disorder with a median life expectancy of nowadays more than 60 years [7, 10].

During the last decade CFTR modulators have arrived at the clinic [11] that target the basic defect in CF of impaired epithelial conductance for chloride and bicarbonate [12–14]. There are two classes of CFTR modulators: Potentiators [15–17] increase the activity of CFTR to the cell surface and correctors facilitate the translation, folding, maturation and trafficking of mutant CFTR to the cell surface and/or prevent its premature degradation [18, 19]. By 2019 a highly effective triple combination of the potentiator ivacaftor (IVA) [15] and the two correctors elexacaftor (ELX) [20] and tezacaftor (TEZ) [21, 22] have become available for the treatment of the more than 80% of pwCF who harbor at least one CFTR allele that is responsive to this medication [23–25]. ELX/TEZ/IVA is the first CFTR modulator therapy shown to halt lung function decline over an extended time period [26, 27]. Based on an individual person-level microsimulation model the median lifetime survival of pwCF receiving triple modulator therapy and best supportive care has been estimated to be 72 years [28].

Although CF is a rare disease, its advances in diagnosis, therapy and management have it made a showcase for the potential achievements of modern medicine reflected by more than 65,000 entries in the PubMed database by the end of 2024.

Main text

Cystic fibrosis – a monogenic disease personalized by hundreds of disease-causing mutations in the CFTR gene

CF is caused by two mutations in trans in the CFTR gene [5]. As of the most recent file (25 September 2024), 1,167 sequence variants are annotated on the CFTR2 website, 1,085 of which were classified as disease-causing [29]. By December 2024, CFTR France had classified 989 sequence variants, about 80% of which are nucleotide substitutions and the remaining 20% are insertions, deletions, repeats or rearrangements of the CFTR sequence [30]. The DNA alterations lead either to amino acid substitutions (67.7%) or to premature stop codons (14.8%), frameshifts of the reading frame (12.6%), gain or loss of amino acids (1.3%) or to no change of the reading frame (3.6%). Of the 989 sequence variants, 500 and 76 variants were discerned as disease-causing and non disease-causing (‘benign polymorphisms’), respectively [31]. By combining in silico predictions and wet lab, clinical and epidemiology data, the remaining sequence variants were assigned to the intermediate categories of ‘likely benign’ (0.7%), ‘likely pathogenic’ (4.6%) or variants of unknown significance (VUS, 36.5%) [31]. VUS carriers may be clinically asymptomatic or may present the features of mild CF or CFTR-related disorder (CFTR-RD) [32, 33] such as congenital absence of the vas deferens (CAVD), diffuse bronchiectasis, chronic rhinosinusitis, chronic or acute recurrent pancreatitis, allergic bronchopulmonary aspergillosis (ABPA), primary sclerosing cholangitis and aquagenic wrinkling [34, 35]. These eight conditions typically have an etiology unrelated to CFTR. Since CFTR genetics is not informative, any suspect case of CFTR-RD has to be examined by CFTR biomarkers (see below) to make a diagnosis [33].

The most common disease-causing CFTR mutation, the 3-bp in frame deletion F508del (c.1521_1523 delCTT; p.Phe508 del according to the HGVS Nomenclature [36]) accounts worldwide for approximately 60—70% of all CF chromosomes with variable frequencies depending on populations and geographical locations [5]. Within Europe, the frequency varies from 32% in Turkey to 83% in Denmark [5, 37]. The majority of other mutations is rare. For example, in Germany just eight disease-causing CFTR mutations are present at a frequency of more than 1% in the population of CF alleles [10].

Molecular pathology of CFTR mutations

The ABC transporter CFTR (ABCC7) is a low-conductance anion-selective ion channel with complex regulation [38–40], namely CFTR activation by protein kinase dependent phosphorylation and CFTR channel gating by ATP binding and hydrolysis [39–41]. CFTR is assembled from 1,480 amino acids organized into five major domains, i.e. two membrane-spanning domains (MSDs), two nucleotide-binding domains (NBDs) and a regulatory domain (RD) arranged in two MSD-NBD motifs separated by the RD [38, 42, 43]. The MSDs are composed of six transmembrane segments (TMs), which assemble to form a gated pore for transmembrane anion flow [44, 45]. The pore is accessible from the cytosol through a large inner vestibule and opens to the extracellular solvent through a narrow portal made up by the extracellular ends of TMs 1, 6, and 8 [46].

CFTR transports chloride to control fluid absorption or secretion by epithelia and conducts bicarbonate to regulate intra- and extracellular pH [12–14]. Depending on the amount of residual CFTR activity conferred by the CFTR mutation genotype, CFTR function may be absent, modified and/or reduced in pwCF. Based on functional criteria related to chloride transport, CFTR mutations have been assigned to six classes [47]:

• Class I (minimal function mutations) includes structural variants such as large deletions or rearrangements and sequence variants that result in a premature termination codon (PTC) or cause a shift in the reading frame.

• Class II mutations lead to defective posttranslational maturation and/or trafficking so that no or only small amounts of mutant CFTR are translocated to the apical membrane of epithelial cells.

• Class III mutants are defective in the regulation and gating of the ion channel.

• Class IV mutants have altered channel properties (open probability, open time, ion selectivity, conductance).

• Class V mutants are characterized by a reduced amount of normal CFTR protein.

• Class VI mutants have a reduced half-life leading to the absence or severe reduction of CFTR at the membrane surface.

Patients who harbor two class 1, 2 or 3 CFTR alleles typically suffer from the whole spectrum of CF disease including exocrine pancreatic insufficiency (PI) [48]. Individuals with two CFTR mutations in trans who carry at least one class 4 or 5 CFTR allele are affected by milder forms of CFTR-associated disease; i.e. CFTR-RD or CF with exocrine pancreatic sufficiency (PS) [48].

Diversity of patients’ basic defect assessed by CFTR biomarkers

The CFTR mutation genotype translates into defects that disrupt CFTR production, channel activity and/or regulatory function. CF is a systemic disorder of exocrine epithelia. However, in CF patients the investigation of the basic defect is currently limited to few organs and cell types, namely the sweat gland, the respiratory, renal and intestinal epithelium and host defense cells [33]. The repertoire may be increased in the next years by protocols that differentiate CF patient-derived induced pluripotent stem cell lines into cell types of interest [49]. Here we describe the currently available CFTR biomarkers sorted by organ.

Sweat gland: CFTR-mediated secretion in the coil and chloride reabsorption from the duct

The sweat gland provides two biomarkers of CFTR activity: a linear readout via the ß-adrenergic sweat rate [50] and a logarithmic readout via sweat chloride concentration [51]. Sato and Sato [52] as well as Behm et al. [53] discovered in the 1980 s that CFTR-mediated sweat secretion in the secretory coil can be stimulated with ß-adrenergic agonists if the thermoregulative cholinergic sweat secretion is simultaneously inhibited with atropine. Current ß-adrenergic sweat secretion assays measure the sweat gland secretory responses to this stimulation sequence by either evaporimetry [54, 55] or by recording the growth of single droplets under a microscope [50, 55–57]. Protocols have recently been developed that determine sweat secretion rates by automatic recording, processing and quality control of the kinetics of sweat droplet formation [56, 57]. The ß-adrenergic sweat rate increases approximately linear with CFTR function. The sweat rate is close to 50% in heterozygous carriers of a CFTR mutation (median: 0.26 nl/min) compared to healthy controls (median: 0.44 nl/min) and is absent or barely detectable in pwCF (median: 0.006 nl/min) [57]. Thus, the ß-adrenergic sweat secretory test is the first CFTR biomarker assay indicating half-maximal CFTR function in heterozygous gene carriers. It also identifies all pwCF including those carrying mutations associated with normal or borderline sweat chloride concentrations in the quantitative pilocarpine iontophoresis sweat test (QPIT) [58].

QPIT [58, 59] assesses the CFTR-mediated reabsorption of chloride from the primary plasma isotonic sweat in the sweat duct. Based on the abnormally low chloride permeability of the CF sweat duct [60, 61], QPIT is the standard CFTR biomarker in making or excluding a diagnosis of CF. After cholinergic sweat stimulation with pilocarpine iontophoresis, sweat is collected and its chloride concentration determined. Sweat chloride concentration (SCC) varies between 5 and 140 mmol/L [62], whereby 60 mmol/l is the cutoff for a CF diagnosis [63]. CFTR activity decreases with the logarithm of SCC [51].

Data of the CFF Patient registry collected from 25,753 patients between 1996 and 2009 uncovered an association between the SCC at diagnosis and survival at the population level [64]. Median survival age was significantly different between the three sweat chloride categories (SCC < 60 mmol/L, SCC 60–80 mmol/, SCC > 80 mmol/L), with the < 60 mmol/L group having the highest median survival age, followed by the 60–80 mmol/L group [64]. Likewise, median age of survival was significantly lower for patients with two class I, II or III CFTR mutations compared to pwCF carrying at least one class IV or V CFTR mutation [64]. Data from twin and sibling studies demonstrated that CFTR mutations are the primary determinant of SCC variability (56% of variation) followed by variation over time (age, circadian cycle, day-to-day variation) and environmental factors [62]. The intrapair comparison of SCC of monozygous and dizygous twins suggested that genetic modifiers do not significantly influence the outcome of sweat testing [62]. These findings corroborate the clinical practice that QPIT is the adequate test to diagnose CF.

The SCC of F508del homozygotes in normally distributed around a mean of 101 ± 15 mMol/L [62]. During his 40-year working life, the first author met a few F508del homozygous subjects with a SCC below 70 mmol/L. They all presented at the CF clinic with extremely mild CF disease and a late diagnosis in their 20 s to 70 s in line with the population data that a low SCC is associated with higher survival.

CFTR-mediated chloride and bicarbonate homeostasis in airway epithelia and host defense cells

The clinical manifestation of CF is primarily determined by pulmonary, hepatobiliary and intestinal disease that is triggered by the CFTR mutation type, but then is predominantly shaped by genetic modifiers and environmental factors [65, 66]. Within the lungs, CFTR regulates the pH and hydration of the airway surface liquid (ASL) [67]. The tasks are distributed to different cell types. The abundant secretory and basal cells secrete chloride [68]. On the other hand, the rare CFTR-rich ionocytes [69, 70] mediate the absorption of chloride and fluid from the ASL [71] and modulate its pH by secreting bicarbonate via CFTR-linked chloride/bicarbonate exchange [72]. These processes can be followed in samples from individual CF patients of any age by taking nasal swabs for immediate omics analysis [73] or set-up of cell cultures [71, 74, 75]. After the initial expansion in co-cultures, the epithelial cells are grown at the air–liquid interface for subsequent biochemical, histological or electrophysiological characterization [75, 76]. The nasal brushings harvest the whole spectrum of ionocytes, basal, club, goblet and ciliated cells so that the patient’s basic defect can be assessed at the level of single cells in a real-world scenario [73]. Moreover, one gets access to the host defense cells [73] including the macrophages that are known to be defective in CFTR-mediated phagosome acidification [77] undermining their intracellular capability to lyse CF pathogens such as Pseudomonas aeruginosa [78–80].

Besides these novel CFTR biomarkers at the cellular level, nasal transepithelial potential difference (NPD) measurements represent the CFTR biomarker of the respiratory epithelium at the macroscopic scale [81]. NPD measures the chloride conductance of the nasal respiratory epithelium in vivo. NPD tracings of pwCF are characterized by a more negative basic potential, a higher hyperpolarization potential upon amiloride-induced blockage of sodium conductance and an absent or low depolarization potential upon exposure to chloride-free solution because the CFTR-mediated chloride conductance is low or absent [82–84]. Residual CFTR activity of patients carrying a class IV or V mutation shows up in a small depolarization potential [85, 86]. To differentiate CF from non-CF, NPD scores have been developed that evaluate both the hyper- and the depolarization potential [86, 87].

Functional and immunochemical analysis of CFTR in rectal biopsies and organoids

Intestinal current measurements (ICM) of rectal biopsies in Mini-Ussing chambers represent the CFTR biomarker that assesses CFTR-mediated chloride secretion in the colonic epithelium [88–90]. CFTR is the dominant chloride channel in human colon responsible for chloride and fluid secretion. ICM evaluates the chloride secretory responses to cAMP-activation via forskolin/IBMX and cholinergic co-activation via carbachol that are mostly driven by CFTR in healthy non-CF tissue [91, 92]. However, in CF with no or low amounts of functional CFTR the anion secretory activities are substantially mediated by bicarbonate through bestrophin Best2 channels [93], by the Epac-Rap-PLC-[Ca²⁺] signaling pathway [94] and increased by basolateral and apical potassium channels [95, 96]. Thus, to discern CFTR-mediated chloride secretion from other anion secretory activities, one may add at the end of the ICM experiment either the specific CFTR inhibitor CFTR(inh)− 172 [97] or re-stimulate with histamine at a high concentration of DIDS that inhibits all anion channels but CFTR [89, 92]. As in NPD residual CFTR function is typically detected in pwCF who are carrying a class IV or class V mutation on at least one CF allele [98, 99].

The rectal biopsies can be utilized for immunochemical CFTR protein analysis. Immunoblots visualize the CFTR glyco-isoforms. The polypeptide chain is synthesized in the endoplasmic reticulum (ER) (CFTR band A). Its mannose-rich core-glycosylated ER form (CFTR band B) is translocated to the Golgi apparatus for glycan processing before it can reach the apical membrane as complex glycosylated protein (CFTR band C) [100, 101]. The major CF-causing mutation F508del is a class II mutation defective in posttranslational processing and trafficking [101, 102]. Rectal biopsies of F508del homozygous donors contained the F508 del CFTR glycoform C in amounts of 0% to 14% (median: 3%) of that of wild-type CFTR of non-CF donors [103]. Apart from the colon, apical complex-glycosylated F508 del CFTR has also been detected in patients’ airways, gallbladder and intestine, but not in the sweat gland [104, 105].

ICM has to be performed on freshly excised biopsies. Alternatively, the biopsies may be converted to self-renewing and self-organizing organoids that closely resemble the in vivo tissue architecture in terms of cell polarity, self-renewal kinetics and cell-type composition [106, 107]. The organoids contain a central lumen lined by the apical membrane, and multiple crypt domains that harbor the stem cells. A non-CF rectal organoid is filled with fluid and has a cystic, spherical appearance [108]. Raising cAMP concentrations by forskolin results in CFTR channel opening and rapid swelling [49]. Forskolin-induced swelling (FIS) is completely CFTR dependent [49]. CF rectal organoids show strongly reduced or absent forskolin-induced swelling, a lower lumen volume and a more dense and irregular structure [49, 109, 110]. Both the FIS assay and the rectal organoid morphology analysis (ROMA) have become CFTR biomarkers applicable to diagnosis and personalized treatment [49, 110]. In contrast to colonic or rectal organoids, non-CFTR dependent organoid swelling can be observed in organoids derived from airway, kidney, pancreas, bile ducts and epididymis, and swelling can be induced by other agonists [111, 112]. Thus, the easily accessible rectal organoids are particularly suited to assay the CFTR function of individual CF patients.

CFTR-dependent renal bicarbonate secretion

The most recently introduced in vivo biomarker of CFTR function is the challenged urine bicarbonate test [113, 114]. The chloride-secreting CFTR channel and the chloride/bicarbonate exchanger pendrin (SLC26A4) are co-localized in the ß intercalated cells of collecting ducts of the kidney [115]. During acute systemic base excess urine secretion of bicarbonate is increased in a CFTR-dependent manner [116], which is examined in the test: First, urine is collected at baseline, and subsequently, each participant ingests 79 mg/kg body weight (corresponding to 0.94 mmol/kg body weight) of NaHCO₃ dissolved in 200 mL of tap water. Urine is collected hourly for 3 h and 200 mL of tap water is ingested after each urine output for the first 2 h. The functional readout of the test is the total amount of bicarbonate excreted during the 3 h after ingestion of NaHCO₃ [113, 114]. Bicarbonate secretion was significantly decreased in pwCF with class I or class II mutations, but was not significantly different from healthy controls in pwCF who are harboring a class IV or class V mutation on one CF allele [113, 114].

CFTR-mediated interorgan metabolite exchange

Recently, arteriovenous metabolomics in CF newborn pigs have revealed that CFTR is regulating multiorgan metabolism [117]. In CF piglets, the number of metabolites exchanged between the liver and other organs decreased by half compared to that of their wild type littermates (from 140 to 68 metabolites). The number of metabolites transferred to the lung from other organs decreased by fivefold (from 68 to 13 metabolites) whereby particularly the uptake of long chain fatty acids was severely impaired. CFTR loss moreover disrupted the liver release of metabolites and the liver-muscle exchange of glutamate and glutamine. CF kidneys exhibited insufficient reabsorption of many amino acids presumably because the lack of a CFTR anion conductance may change the driving force for sodium and proton-coupled amino acid absorption in the renal proximal tubule. Loss of CFTR also disrupted renal glucose homeostasis because gluconeogenic substrates such as lactate and amino acids were poorly reabsorbed in the proximal tubule. These spectacular, novel findings of the endogenous role of CFTR in interorgan metabolite exchange need to be verified in further work including clinical studies in pwCF until (selected) metabolites will become the next generation of CFTR biomarkers.

Action of CFTR modulators on mutant CFTR

About 90% of pwCF are homozygous or compound heterozygous for the most common mutation F508del [2–5, 65]. F508del CFTR is defective in posttranslational processing and trafficking [19, 101, 102]. Newly synthesized F508del CFTR fails to adopt a wild-type fold in the endoplasmic reticulum (ER), is targeted to ER-associated degradation and is removed faster from the apical membrane by endocytosis [101, 102]. Consequently, F508del CFTR confers no or low chloride and bicarbonate secretory activity. Correction of the conformational defects of F508del CFTR requires the stabilization of the interfaces between the two NBDs and MSDs (type I corrector) and the stabilization of NBD2 (type II corrector) and F508del NBD1 (type III corrector) [19].

The CFTR modulators lumacaftor (LUM) [18, 118] and tezacaftor (TEZ) [21, 22, 118] are type I correctors [19]. The two drugs stabilize the early steps of F508del CFTR biogenesis at the ER [119, 120], facilitate the subsequent domain assembly in the absence of folded F508del-NBD1 and stabilize the conformation of the folded protein [118–128]. The type III correctors elexacaftor (ELX) [23, 24] and vanzacaftor (VAN) [129] synergistically restore the processing and stabilize the conformation of F508del CFTR in combination with type I or type II correctors [130–134].

The CFTR potentiators ivacaftor (IVA) [15] and deutivacaftor (DVA) [129] enhance channel activity by increasing pore opening while NBDs are dimerized [43]. Channel opening normally requires the binding and subsequent hydrolysis of ATP [40, 41, 135]. Ivacaftor reversibly enhances ATP-independent opening of the channel [136–138] by stabilizing pre-hydrolytic states [139, 140] and thereby improves ion transport in F508del CFTR and overcomes the defective ATP-dependent opening of CF-causing gating mutations [141, 142]. In addition to being a corrector, elexacaftor also act as a co-potentiator of F508del, G551D and M1101K CFTR chloride channels [143–145].

Cryo-electron microscopy of reconstituted recombinant protein revealed that the conformations of wild type CFTR and ELX/TEZ/IVA-bound F508del CFTR were almost indistinguishable from each other indicating that upon binding of the three CFTR modulators F508del CFTR is converted into a wild type conformation [146].

CFTR biomarker response to triple CFTR modulator therapy with ELX/TEZ/IVA in pwCF harboring one or two F508del alleles

The advent of the triple combination therapy with ELX/TEZ/IVA has been qualified as a game changer for CF [147]. By the end of 2024, nine phase 3 clinical trials [23–25, 148–152] and twenty real-world post-approval studies [26, 27, 153–170] had consistently reported significant improvements of anthropometry (BMI: median 1.3, IQR 1.1—1.6 kg/m²) and lung function (ppFEV1: median 11.4, IQR 9.8—13.7%) (Table 1). ELX/TEZ/IVA improved lung ventilation and abnormalities in lung morphology [168, 171–181], including airway mucus plugging, wall thickening and in a few cases even bronchial dilatation [182] and bronchial artery dilatation [183]. Microbial load of the airways with CF pathogens such as Aspergillus fumigatus [184, 185], Staphylococcus aureus [186], Pseudomonas aeruginosa [186–188] or nontuberculous mycobacteria [189–191] was reduced, but persisted in most pwCF [186, 192–194] and the microbial network remained vulnerable to fragmentation [195–197]. In the gut, the transit time of food through the small bowel increased [198], but gastrointestinal symptoms remained prevalent [198, 199] or improved only slightly [200–202]. Weight gain was mainly caused by an increased fat mass [203, 204] associated with the risk to develop obesity and metabolic syndrome [204, 205].

Table 1.

Change of CFQ-R score, ppFEV1, BMI and SCC in pwCF after initiation of triple EXL/TEZ/IVA therapy [23–27, 149–171]

^#NL The Netherlands; *BMI-z-score; BMI Body mass index, CFQ-R Cystic fibrosis questionnaire – revised application, F F508del, IVA Ivacaftor, MF Minimal function mutation, ppFEV1 Percentage of predicted forced expiratory volume in 1 s, pwCF people with CF, RF Residual function mutation, SCC Sweat chloride concentration, TEZ Tezacaftor

Of the CFTR biomarkers, sweat chloride concentration (SCC) as a surrogate of CFTR activity had commonly been implemented in post-approval studies and as secondary endpoint in the phase 3 trials (Table 1). Upon initiation of therapy with ELX/TEZ/IVA, SCC significantly decreased in adolescents and adults who are homozygous for F508del or compound heterozygous for F508del and a class I mutation. The median value of the median of 21 clinical studies was – 49.0 mmol/L (IQR − 45.0—− 53.4 mmol/L) (Table 1). The decrease of SCC was larger in children [206] and smaller in pwCF who were carrying at least one class IV or class V mutation associated with lower SCC at baseline [151].

SCC shows a log-linear association with CFTR activity [51, 207]. Pooled data of the phase 3 trials with IVA, TEZ/IVA or ELX/TEZ/IVA demonstrated a mean absolute improvement of 7.3%, 16.9%, 19.6% and 20.9% in ppFEV1 in study participants showing a SCC ≥ 80 mmol/L, ≥ 60—≤ 80 mmol/L, > 30—< 60, < 30 mmol/L in sweat test after 24 weeks of modulator therapy, respectively [207]. Likewise, the three groups showed a differential gain of BMI at week 24 of 0.25, 1.02, 1.16 and 1.27 kg/m² [207]. Thus, greater reductions in sweat chloride were associated with improved clinical outcomes. The PROMISE post-approval study [27] confirmed this finding. The change in sweat chloride from baseline to six months after ELX/TEZ/IVA initiation significantly correlated with the improvement in ppFEV1. Linear regression of the SCC data indicated that a decrease of a 10 mM increment in SCC was associated with a mean 0.89 greater change in ppFEV1 [27].

In contrast to the uniform median response of SCC to ELX/TEZ/IVA therapy at the population level, the individual response was heterogeneous [208]. The US CHEC-SC Study group has recently published SCC data from 1,769 pwCF receiving ELX/TEZ/IVA, of whom 56.6% were homozygous for F508del, 24.7% compound heterozygous for F508del and a class I mutation and the remaining 18.7% were carrying other CFTR genotypes included in the FDA label [208]. 127 pwCF (7.3%) remained in the classical PI CF range with SCC of more than 80 mmol/L; 243 pwCF (13.7%) had a SCC of 60–80 mmol/L; 939 pwCF (53.1%) had a SSC in the borderline range of 30–60 mmol/L and 460 pwCF (26%) had a SCC in the range of healthy people below 30 mmol/L [208]. Thus, close to 80% of pwCF receiving ELX/TEZ/IVA had crossed the threshold of 60 mmol/l into the non-CF diagnostic range. Conversely, 18% of F508del homozygotes remained in the CF range with SCC ≥ 60 mmol/l compared to 31% of compound heterozygotes for F508 del and a class I mutation. Within subgroups sharing the same CFTR genotype, substantial variability of post-ELX/TEZ/IVA SCC was observed in the CHEC-SC study cohort. The US experience was confirmed in CF centers in Europe. For example, the six-month changes from baseline in SCC ranged from − 27.1 to − 76.1 mM for the 10 th to 90 th centile in 211 pwCF seen at the CF center in Milan [163]. In this single center study, the change in SCC weakly correlated with absolute changes of ppFEV1, but did not correlate with the change in BMI [163]. In summary, treatment with ELX/TET/IVA led to a substantial and highly significant decrease of SCC in sweat test at the population level (Table 1), but the response of the individual patient was highly variable. Age, sex, ethnicity and pre-modulator SCC explained less than 10% of the variability in post-ELX/TEZ/IVA SCC in pwCF with the same CFTR genotype [208].

The authors’ consortium ‘Modulate-CF’ in the German Center of Lung Research (DZL) examined the effects of ELX/TEZ/IVA on CFTR function in pwCF with one or two F508del alleles in three biomarkers, i.e. QPIT, NPD and ICM [157]. The median decrease of SCC of 48.5 mmol/L matched with that of the pooled data of the 21 studies (Table 1). 8–16 weeks after initiation of ELX/TEZ/IVA, CFTR function improved in nasal respiratory and intestinal epithelia to a median level close to half of normal [157]. However, the responses of the individual patients were highly variable. The Venn diagram in Fig. 1 shows the numbers of patients who developed a normal response in none, one, two or three CFTR biomarkers. Seven of the 79 study participants remained in the CF range for all three biomarkers and again seven patients achieved CFTR activity in the normal range for all three biomarkers. 33 and 32 study participants normalized in one or two biomarkers, respectively (Fig. 1). We were curious to know how these mixed responses in the basic defect of sweat gland, airway and intestine were associated with the clinical outcome in lung function and anthropometry, the key endpoints of the clinical trials. Based on registry data of European CF populations in the pre-modulator era as reference [209, 210], the FEV1 and BMI data of each patient were converted into age- and gender-corrected disease-specific percentiles [211]. The paired CF percentiles at baseline and after initiation of ELX/TEZ/IVA were mapped for each study participant on the two-dimensional percentile plot shown in Fig. 2. Normalization of CFTR biomarkers is indicated by colored symbols. More than 90% of study participants improved their CF percentile during the first 8 to 16 weeks of triple modulator therapy irrespectively of whether they were affected from mild or severe CF disease (Fig. 2). The normalization of SCC, NPD and/or ICM scattered over the whole range of disease percentiles (Fig. 2). In other words, we observed personal signatures of the normalization of the basic defect in sweat gland, airway and intestine that were not associated with the severity of CF disease at baseline and its improvement during the first weeks of treatment with ELX/TEZ/IVA. Restoration of CFTR function was diverse and not predictive for clinical outcome in the individual patient. For example, the patient with the largest absolute gain of ppFEV1 in our cohort exhibited the smallest change in sweat chloride. In summary, the individual responses of basic defect and clinical outcome measures were heterogeneous upon initiation of ELX/TEZ/IVA.

Fig. 1.

Venn diagram of the response of CFTR biomarkers of pwCF with one or two F508del alleles to 8–16 weeks of treatment with ELX/TEZ/IVA compared to baseline. The Venn diagram indicates the number of participants who normalized the response to the respective biomarker. A study participant showed a normalized response in the respective biomarker if SCC was below 60 mmol/L in QPIT [63], NPD had revealed a Sermet score of more than 0.27 [87], and the chloride secretory responses in the ICM had reached 70% of the mean normal values for at least two of three criteria, i.e. the response to activation with forskolin/IBMX, response to carbachol or the sum of the two responses [85, 91, 92]. Of the 79 study participants, 52 pwCF had been modulator naïve at baseline and 27 pwCF were administered combination therapy with tezacaftor and ivacaftor for at least six months. The numbers in brackets indicate modulator-naïve (left) and TEZ/IVA-positive pwCF at baseline (right) (Source: The authors’ Modulate-CF DZL consortium Berlin, Gießen, Hannover, Heidelberg [157])

Fig. 2.

Personal clinical outcome and response to CFTR biomarkers of pwCF with one or two F508del alleles to 8–16 weeks of treatment with ELX/TEZ/IVA compared to baseline. Patients’ BMI and FEV1 values were mapped onto the age- and gender-corrected percentile distribution of the 2013 edition of the CF European registry [209]. Next, the FEV1 and BMI percentiles were combined into the CF disease percentile, which gives equal weight to lung function and anthropometry [211]. The figure shows for each of the 79 study participants the coordinate of CF disease percentiles at baseline and after 8–16 weeks exposure to ELX/TEZ/IVA. The diagonal separates individuals with improved or worsened percentile ranks upon initiation of ELX/TEZ/IVA. The individual’s normalization of CFTR biomarkers (defined as in Fig. 1) is indicated by symbol: asterisk, no normalization; red triangle, solely SCC [63]; blue triangle: solely NPD Sermet score [87]; yellow triangle, solely ICM [85, 91, 92]; purple square, SCC and NPD Sermet score; green square, NPD Sermet score and ICM; brown square; SCC and ICM; open circle, SCC + NPD Sermet score + ICM (Source: The authors’ Modulate-CF DZL consortium Berlin, Gießen, Hannover, Heidelberg [157])

Despite being diverse, almost all responses to ELX/TEZ/IVA in the DZL cohort reflected improvements of CFTR function and general condition. Thus, we were taken by surprise to learn that this positive trend was not replicated by the β-adrenergic sweat secretion assay in a subgroup of the DZL cohort [57]. Triple therapy with ELX/TEZ/IVA normalized SCC, but gained only 4% of median wild-type β-adrenergic sweat rate [57]. β-adrenergic sweat stimulation in the coil is apparently more stringent in its requirement for CFTR activity than chloride reabsorption in the duct measured by the sweat test, because CFTR is only a minor component in the secretory coil and becomes the rate-limiting step of chloride secretion upon ß-adrenergic stimulation [52]. On the other hand, CFTR-mediated bicarbonate secretion in the kidney is more tolerant to compromised CFTR function than CFTR-mediated secretion and reabsorption of chloride in the sweat gland [116]. Bicarbonate excretion was close to normal for pwCF carrying F508del and a residual function mutation and treatment of F508del homozygotes with ELX/TEZ/IVA increased bicarbonate excretion to about 70% of healthy controls [113, 114].

Diversity of the response to ELX/TEZ/IVA in pwCF with one or two F508del alleles should reflect the impact of genetic and environmental modifiers on clinical outcome. F508del CFTR is a class II mutant defective in posttranslational processing and trafficking [19, 101, 102]. The members of the F508del interactome are thus the prime candidates for modifiers of corrector-mediated rescue of F508del CFTR. The amount of the thermolabile F508del CFTR protein in the apical epithelial membrane varies among pwCF and their organs and cells [101–105] and thus is was not a surprise to learn that the ELX/TEZ/IVA-mediated rescue of processing and targeting of F508del CFTR in the rectal epithelium differed substantively among the study participants of the DZL cohort [212]. Likewise, the rectal organoids of F508del homozygous donors showed personalized transcriptome profiles in presence and absence of ELX/TEZ/IVA [213]. Correspondingly, the profile of the top differentially expressed genes in the nasal epithelial transcriptome predicted with about 85% accuracy the individual patient’s improvement of FEV1 and BMI upon ELX/TEZ/IVA initiation [214]. The BIH consortium collected nasal swabs from children with CF and at least one F508del allele aged 6–11 years at baseline and three months after initiation of ELX/TEZ/IVA [73]. Compared to healthy children, CFTR-positive cells were decreased in epithelial basal, club and goblet cells and were restored by ELX/TEZ/IVA therapy to nearly healthy levels. Single cell cDNA sequencing revealed that ELX/TEZ/IVA partially restored the impaired innate immune response in epithelial cells and reduced the pro-inflammatory status of immune cells in the CF nasal mucosa [73]. This scRNA-seq study on airway epithelial and immune cells demonstrated for the first time that highly efficient triple CFTR modulator therapy might restore epithelial homeostasis and host defense in CF airways, particularly if treatment is started in children with preserved lung function [73].

CFTR biomarker response to non-F508del CFTR genotypes

EMA had approved the therapy with ELX/TEZ/IVA for the 90% of pwCF in Europe aged 2 years or more who are harboring at least one F508del allele. Besides F508del, screening of 605 CFTR sequence variants in recombinant Fischer Rat Thyroid (FRT) cells identified further 496 variants (82%) that respond with more than 10% normal CFTR chloride transport activity to ELX/TEZ/IVA in vitro [215]. Thus, ELX/TEZ/IVA treatment should result in clinical benefit for almost all pwCF who do not carry two loss-of-function class I mutations. In the USA, the FDA has already approved 177 of these 605 variants for ELXTEZ/IVA therapy targeting 5.2% of the CF population in the US patient registry [216]. Prior to the expansion of the label in April 2025 by EMA (see below), off-label prescription of ELX/TEZ/IVA for patients with non-F508del CFTR genotypes in Europe has depended on case-by-case negotiations between CF physician and health insurance company. To identify the individuals who probably will gain clinical benefit from permanent therapy with this costly medication, it has become a common procedure to monitor CFTR activity with biomarkers during an initial pilot phase.

Table 2 provides an overview of published case reports of ‘theratyping’ of patient-derived organoids or epithelial cell cultures [76, 217–229]. Of 98 examined specimens from CF donors, 14 samples showed no extra CFTR activity upon exposure to ELX/TEZ/IVA and 61 samples showed a strong response to ELX/TEZ/IVA in the healthy range or in the range observed for CF donors with one or two F508alleles [217, 218]. The remaining 23 samples exhibited CFTR activity that was below the range for the F508del reference. Thus, 85% of samples from pwCF with non-F508del CFTR genotypes were responsive to triple modulator therapy.

Table 2.

Theratyping: CFTR biomarker response of non-F508 del CFTR genotypes to ELX/TEZ/IVA

A. Rectal organoids

a. FIS assay [217–225]

• Response of genotype in the range of healthy or F508 del/F508 del or F508 del/class I mutation genotypes

G85E/A561E; G85E/I1234 V; G85E/W1282X; G85E/N1303K; G85E/N1303K; G85E/CFTRdele17a- 18; G85E/1677delTA; E92K/E92K; R117L;L997F/[R117L;L997F]; Q220X/A1006E; Q220X/A1006E; Q220X/R1066C; R334W/R764X; R334W/N1303K; [R334W;Q378X]/[R334W;Q378X]; R347P/CFTRdele17a- 18; G461E/N1303K; [L467F;F508del]/F508del; G542X/711 + 3 A > G; A559T/A559T; Q1012P/N1303K; R1066C/N1303K; R1066C/R1066H; R1066H/CFTRdele2,3; L1077P/N1303K; S1159F/S1159F; W1282X/N1303K; N1303K/N1303K; N1303K/N1303K; N1303K/2143delT; N1303K/2143delT; N1303K/3821delT; N1303K/3121 - 1 A > G; N1303K/3849 + 10kbC > T; 991del5/3849 + 10kbC > T; 3849 + 10kbC > T/3849 + 10kbC > T

• Response of genotype in the intermediate range

G85E/N1303K; R553X/3272–26 A > G; E585X/N1303K; N1303K/G542X; N1303K/W1282X; N1303K/W1282X; N1303K/N1303K; N1303K/N1303K; N1303K/N1303K; N1303K/N1303K; N1303K/N1303K; N1303K/N1303K; N1303K/4010delTATT; L1335P/L1335P; 1898 + 5G > T/3272 - 26 A > G; 2043delG/4382delA

• No response of genotype

M1V/N1303K; E60X/4015delATTT; R347P/L571S; G550X/N1303K; L927P/W1282X; W1282X/N1303K; N1303K/2184insA; 711 + 1G > T/2789 + 5G > A

b. Ussing chamber measurements of chloride secretion in organoid-derived rectal epithelial monolayers [219, 222, 225, 226]

• Response of genotype in the range of healthy or F508del/F508del or F508del/class I mutation genotypes

A559T/A559T; S737F/W1282X; S737F/CFTRdele22 - 24

• Response of genotype in the intermediate range R347P/R347P; T465N/Q39X

• No response of genotype L227R/L227R

c. Western immunoblot of CFTR protein [219, 222, 225, 226] Band C present: T465 N/Q39X; A559T/A559T; W57G/A234D; R347P/R347P

B. Nasal brushings

a. FIS assay [222, 227]

• Response of genotype in the range of healthy or F508del/F508del or F508del/class I mutation genotypes

W57G/A234D; L1077P/L1077P; L1077P/W1282X; L1077P/R1066 C

b. Ussing chamber measurements of chloride secretion in airway-liquid interface cultures [76, 219, 222, 223, 226, 227, 228, 229]

• Response of genotype in the range of healthy or F508del/F508del or F508del/class I mutation genotypes

L1077P/L1077P; L1077P/W1282X; L1077P/R1066C; N1303K/N1303K; W57G/A234D

• Response of genotype in the intermediate range

G85E/G85E; G1244E/G1244E*; G1244E/1717–1 G-A*; G1244E/G542X*; N1303K/N1303K

• No response of genotype

[L467F;F508del]/G542X; [L467F;F508del]/E585X; W1282X/W1282X; 2184 AA > G/3892delTT

*poor response to potentiator IVA, but strong response to potentiator apigenin

c. Western immunoblot of CFTR protein [226–228]

Band C absent: W1282X/W1282X; Band C present: W57G/A234D; L1077P/L1077P; L1077P/W1282X; G1244E/G1244E; G1244E/1717-1 G-A; G1244E/G542X

NPD and ICM have been rarely applied to test the effect of ELX/TEZ/IVA on CFTR activity. The authors examined a few patients with these biomarkers. Two individuals who are homozygous for G85E or N1303 K showed a response of 32% and 5% of normal in the NPD and of 29% and 15% in the ICM, respectively [76]. Two brothers who are compound heterozygous for class I donor and acceptor splice mutations that affect the same exon, presented an unusual response to ELX/TEZ/IVA [230]. Both patients showed no improvement in QPIT, NPD and ICM and clinical outcomes, but their CFTR-mediated β-adrenergic sweat rate normalized into the healthy range. To put this surprising finding into perspective, the reader may be reminded that pwCF with one or two F508del alleles will always improve in QPIT, NPD and ICM upon initiation of ELX/TEZ/IVA (see Fig. 1), but only about 10% of pwCF will gain a β-adrenergic sweat rate in the normal range after introduction of triple therapy [57].

The class II mutation N1303K present in 43 of 196 analyzed alleles (Table 2) received considerable attention of the ‘theratyping’ community. Located in the second nucleotide binding domain NBD2, N1303K CFTR is distinct from F508del CFTR in its channel properties and its pathway of posttranslational processing and trafficking [231–236]. Since N1303K CFTR was non-responsive to ELX/TEZ/IVA in recombinant FRT cells in vitro, N1303K was not included into the FDA-approved label. However, theratyping of patient-derived organoids and cell cultures with one or two N1303K alleles uncovered a broad spectrum of no (9%), intermediate (47%) or F508del-like responses (44%) to ELX/TEZ/IVA (Table 2). Based on the in vitro experimental evidence that N1303K CFTR showed activation with ELX/TEZ/IVA in their personal organoids, numerous patients commenced treatment with ELX/TEZ/IVA in the frame of clinical trials [223, 237, 238]. Study participants significantly improved in ppFEV1, CFQ-R and BMI, but showed no decrease in SCC and a variable personal response in NPD. Interestingly, rescue of N1303K and F508del CFTR was found to be enhanced under inflammatory conditions that are typical for CF airways [239, 240]. Sweat glands are not inflamed and thus the responses to CFTR modulators may be more limited than in inflamed CF airways or intestine. The supportive role of inflammation for the clinical efficacy of triple CFTR modulator therapy may also give a hint to understand why “changes in sweat Cl⁻ in individual pwCF are poorly predictive of clinical responses despite robust associations between sweat Cl⁻ responses and clinical efficacy on the population level” (Martina Gentzsch in ref. [239]).

The unanticipated clinical benefit of triple modulator therapy of the FRT-refractory N1303K mutation may not only reflect the broad spectrum of responses of individual patient’s CFTR mutant and its interactome to ELX/TEZ/IVA, but could also arise from off-target effects of the medication. Indeed, tezacaftor has been shown to alter the balance of the de-novo synthesis of sphingolipids by inhibition of sphingolipid delta- 4 desaturase that converts dihydroceramides into ceramides [241, 242]. Ceramide accumulates in the lungs of pwCF causing chronic inflammation, impairment of mucociliary clearance and susceptibility to bacterial infection [243–245]. Treatment with ELX/TEZ/IVA partially normalized the disturbed plasma sphingolipid profile in pwCF [246]. Tezacaftor does not only affect sphingolipid metabolism, but it also inhibits the SarcoEndoplasmic Reticulum Calcium ATPase (SERCA) [247]). Thereby tezacaftor and also elexacaftor normalize calcium homeostasis in F508 del-CFTR cells in a CFTR independent manner [248].

The critical anion secretion defect in CF is caused by diminished apical CFTR channel activity. As CFTR function is restored with the use of CFTR modulators, mechanisms that import chloride and bicarbonate across the basolateral membrane or generate bicarbonate within the cytosol may become rate limiting [240]. Rehman and colleagues now have shown that TNF-α + IL-17 treatment sensitizes primary CF airway epithelia to the beneficial effects of ELX/TEZ/IVA via p38 MAPK signaling [240]. The proinflammatory cytokines TNF-α + IL-17 increased the expression of CFTR and of Cl^– and HCO₃^– importers and improved mucociliary clearance by lowering ASL viscosity.

These extra CFTR-independent benefits probably represent personal signatures. The variability of the transcriptome of F508del-homozygous airway epithelial cells was explained to 70% by differences between donors and only to 6% by the presence or absence of treatment with ELX/TEZ/IVA [249]. This strong Garrodian individuality of the CF airway transcriptome and its response to ELX/TEZ/IVA justify a posteriori the decision of the French Compassionate Program to prescribe ELX/TEZ/IVA for all pwCF with advanced lung disease irrespective of their CFTR genotype [250]. Meanwhile all pwCF without F508del variants who were living in France and aged 6 years or older have become eligible for a 4–6 week trial of ELX/TEZ/IVA [251]. Over half of the 516 treated pwCF responded to ELX/TEZ/IVA [251]. Among 360 participants with no FDA-approved variant and no previous CFTR modulator, 177 (49%) were responders; in responders, mean absolute change in SCC was − 20.5 mmol/L and ppFEV₁ was 13.2 percentage points [251] consistent with the improvement of ppFEV1 seen in pwCF with one or two F508del alleles (see Table 1). CFTR potentiators and CFTR correctors have initially been developed for mutation-specific therapies in CF (reviewed in [247]), but the outcome of the expanded French Compassionate Program [251] tells us that efficient triple therapy with ELX/TEZ/IVA is potentially beneficial for almost all pwCF but the unfortunate ones who are carrying two non-responsive loss-of-function class I mutations. In other words, the Garrodian approach of individual theratyping paved the way to an Oslerian label for almost all CFTR genotypes.

Consistent with this conclusion, the efforts of the academia to show efficacy of ELX/TEZ/IVA for many non-F508del CFTR mutations have convinced the regulators to expand the label. On 04 April 2025, the European Medicines Agency (EMA) adopted extensions to the existing indications for ELX/TEZ/IVA to extend its use to all pwCF aged two years or more who carry at least one non-class I CFTR mutation. In other words, only the small group of pwCF who carry two class I mutations and hence do not produce CFTR protein will remain excluded from triple CFTR modulator therapy.

Future directions

ELX/TEZ/IVA has improved the quality of life of pwCF under real-world conditions. However, the 5 years since approval are too short to conclude whether triple modulator therapy may halt the progression of CF lung disease in the long run. Airway dysbiosis [186, 192, 193, 196], inflammation [252–256] and unpleasant gastrointestinal symptoms [199, 257] persist – caveats we need to bear in mind. Moreover, late effects may emerge during life-long therapy. On the other hand, the younger the patient, the more SCC decreased [150, 179, 206] and the more epithelial homeostasis and host defense were restored [73] upon initiation of ELX/TEZ/IVA suggesting that we can expect maximal benefit of triple modulator therapy if we start treatment early in life in children with preserved lung function. This upcoming generation of pwCF will hopefully experience a rather normal life in childhood and adolescence that is not burdened anymore by extensive therapeutic measures and repeated hospitalization.

Conversely, today’s generation of CF adults who were modulator-naïve for decades, will remain compromised in their health by multiple primary and age-related secondary co-morbidities [7]. Likewise, by early 2025 CFTR modulators are still not a game changer for a large part of the CF patient population. First, access to the costly CFTR modulators is still not in place for eligible pwCF in many countries [258, 259]. Second, we need to establish and optimize dosage regimens for the unlucky few who are affected from relative or absolute contraindications such as severe hepatobiliary disease [260, 261] or treatment of infections with mycobacteria [262]. Third, a subset of patients – termed modulator-refractory CF – continues to experience two or or more pulmonary exacerbations per year requiring hospitalization or intravenous antibiotics, regardless of other modulator benefits [263]. Lastly, an at-risk subgroup of pwCF receiving ELX/TEZ/IVA has been identified [264, 265] who reported negative side effects in their mental well-being, which led to intermittent or definitive discontinuation of drug taking. These adverse neuropsychiatric effects are the most prevalent adverse drug reaction of CFTR modulator therapy [266]. The etiology is unknown. Participants of a workshop organized by the Cystic Fibrosis Foundation recommended that future studies should focus on understanding the role of CFTR in the nervous system, defining ELX/TEZ/IVA impacts in preclinical models [265].

ELX/TEZ/IVA pass the placental barrier and the lactating breast [267]. As demonstrated in the CF ferret, exposure to ELX/TEZ/IVA in utero may prevent fetal and postnatal pathologies associated with CF [268]. Consistent with these findings in the animal model, prenatal ELX/TEZ/IVA through a CF carrier mother and a mother with CF prevented meconium ileus [269] and retained a normal vas deferens in fetuses with CF [270], respectively. To develop ethical principles and consensus guidelines for care of reproductive-aged people on modulator therapy that is not based on accidental case reports, the CF community is urged to collect data in the next years [271] that characterize maternal, fetal and long-term offspring outcomes following CFTR modulator therapy use during pregnancy and breastfeeding [272].

Current CFTR biomarker studies focus on ELX/TEZ/IVA, but this scenario will change in near future. A novel triple therapy (VNZ triple) is available for pwCF in the US since January 2025. The corrector elexacaftor is exchanged by vanzacaftor and the potentiator ivacaftor by its deuterated analogue deutivacaftor [129]. Due to more favorable pharmacokinetics, VNZ triple only needs to be administered once daily. Three phase 3 clinical trials with more than 1,000 study participants tested VNZ triple vs. its comparator ELX/TEZ/IVA for 24 weeks. The absolute change of ppFEV1 from baseline through week 24 was not different between ELX/TEZ/IVA and VNZ triple [273, 274]. VNZ triple was superior to ELX/TEZ/IVA in proportion of participants achieving SCC < 60 mmol/L (86% vs. 77% in pwCF aged 12 years or more and 95% vs. 84% in pwCF aged 6–11 years) and in proportion of participants achieving SCC < 30 mmol/L (31% vs.23% in pwCF aged 12 years or more and 53% vs. 39% in pwCF aged 6–11 years) [273, 274]. Referring to the experience of the differential efficacy of ELX/TEZ/IVA on the N1303K mutant in sweat gland and airways [223, 237–240], the significance of the gain of CFTR activity in the sweat gland on clinical outcomes remains an open question that could be addressed in post-approval studies with more CFTR biomarkers and more sensitive clinical endpoints that had applied by the DZL consortium in the past [157, 171, 179].

So far, all approved CFTR modulators were developed and marketed by Vertex Pharmaceuticals. Other groups from academia and industry have yet not succeeded to bring any of their drugs to approval for treatment of pwCF. However, by 2019 Sionna Therapeutics was formed that has the mission to fully normalize CFTR function. One CFTR potentiator and five CFTR correctors that target locations in MSD2, MSD1, NBD1 or ICL4 are currently being studied in phase 1 or phase 2 clinical trials [275, 276]. By the time of writing, publicly available peer-reviewed reports of preclinical and clinical research with these compounds were still missing.

A further long running option for CF drugs that has not yet being materialized in clinical practice, are compounds that modify sodium or alternative chloride channels [277]. The activation of the Ca²⁺ activated channels TMEM16A [278] and SLC26A9 [279, 280] has been the major focus of research, but divergent modes of action and inapt localization in cells and organs sustain the long-lasting debate whether this CFTR-agnostic approach is beneficial for the treatment of CF [281–283].

The CFTR gene, its cDNA and mRNA transcripts are further targets for personalized therapies of CF [284]. Antisense oligonucleotide (ASO)-based exon skipping for splicing modulation has been developed for the 3849 + 10 kb C-to-T [285] splicing and the W1282X [286–288] nonsense mutations. The ASO drug SPL84 prevents the inclusion of a cryptic exon from the 3849 + 10 kb C-to-T allele and thus leads to an increase of correctly spliced CFTR mRNA and higher levels of functional protein [285]. Inhalation of SPL84 led to a broad distribution in cells and nuclei of mouse and monkey lungs [289] and was safe in healthy volunteers in a phase I study [290].

Besides ASOs, the knockdown of a key player in translation termination, i.e. the eukaryotic release factor 3a [291], and the engineering of suppressor tRNAs are promising strategies to tackle premature stop codons [292, 293]. Albers and colleagues designed a strategy that is based on altering native tRNAs into efficient suppressor tRNAs by individually fine-tuning their sequence to the physico-chemical properties of the amino acid that they carry [293]. The engineered suppressor tRNAs re-established expression and function of CFTR stop mutations in cell systems and CF patient-derived nasal epithelia and restored airway volume homeostasis [293]. Thereby the translation velocity of the sequence upstream of the premature stop codon critically modulated readthrough efficacy [294]. This treatment-response heterogeneity calls for personalized tRNA-based gene therapies in the future [292, 294]. Besides academia, Southern Research, a nonprofit contract research organization, is working on potential therapies for nonsense mutations [276].

The ‘universal strategy’ of gene therapy aims to offer treatment for all pwCF. Wild type CFTR gene, CFTR cDNA, CFTR mRNA or a correction system are delivered to the target of interest. Unless stem cells are manipulated, gene therapy allows no systemic therapy as with modulators, but will target one preferred organ such as the CF lung. Prime editing is the most recent addition to the portfolio of programmable gene-editing tools [295]. In contrast to CRISPR/cas genome correction by gene replacement [296], knock-in [297] or homology-directed repair [298], prime editors do not require the creation of double-strand DNA breaks, minimizing any unwanted on-target and off-target editing [295]. Its potential for curative CF gene editing therapy has recently been demonstrated in cell lines, organoids and primary cells for the L227R, N1303K and F508del mutations [299, 300]. The optimization of prime editing for the mutation of interest is not trivial. The reader is recommended to learn from the pioneers’ report about the multiple steps that were necessary to finally gain functional correction of F508del CFTR in airway epithelium [300].

Pulmonary delivery of nucleic acids to the CF lung epithelia has substantially advanced during the last 30 years and can now be executed by inhalation with lipid nanoparticles [298, 301, 302] or recombinant lentivirus [303, 304] or adeno-associated virus (AAV) [305]. Preclinical research currently focuses on full-length CFTR gene delivery, gene editing, phages as vectors and novel formulations [276]. The on-going phase 1 and phase 2 trials investigate aerosolized delivery of CFTR mRNA in lipid nanoparticles or of CFTR cDNA with AAV or lentivirus vectors [276, 303, 304]. To remain up-to-date of this fast-moving field, the reader may regularly consult the CFF website [276].

The UK Respiratory Gene Therapy Consortium whose scholars probably gained worldwide the most profound expertise for CF gene addition therapy within the last 20 years [306], will execute the first-in-human trial with a recombinant lentiviral vector in pwCF [303, 304, 307]. The third-generation lentiviral vector rSIV.F/HN carries a codon-optimized CpG-depleted CFTR cDNA, is pseudotyped for receptors on the apical surface of airway epithelia and has been engineered to minimize the risk of insertional oncogenesis, the major safety concern of lentiviral vectors [303, 304]. Lentiviral vectors integrate into the genome of transduced cells so that the cDNA insert can be expressed throughout the lifetime of the cell. The lentiviral vectors can incorporate comparably large inserts such as the full-length CFTR cDNA and evoke only minor inflammatory host responses. Based on the UK consortium’s meticulously performed preclinical work on vector design to optimize safety and efficacy, we can expect the maximal possible benefit of gene therapy with a viral vector for pwCF that is feasible with current knowledge and technology.

Conclusions

All nucleic acid – based approaches of the last 35 years to cure CF succeeded to demonstrate the proof of principle in cells and animal models, but failed to correct CF in humans. Despite the tremendous improvements of tools to edit any human gene in the genome with high precision and minimal off-target effects, the formulation, delivery, and transduction of vector and its long-term correctly regulated gene expression remain challenging problems. The CF community will be anxious to learn whether Lenticlair™ of the UK Respiratory Gene Therapy Consortium [303, 304] will make a difference and will bring for the first time mutation-agnostic gene addition therapy to the CF patient. The optimal outcome would be the restoration of CFTR function in the lung mimicking the in vivo distribution of tasks between basal cells, secretory cells and ionocytes to secrete and absorb chloride and bicarbonate [67–72].

However, even the best possible outcome will correct function in just one major organ of the systemic disorder CF. Thus, the success of triple CFTR modulator therapy with small molecules cannot be valued highly enough as it has brought to almost all pwCF the partial restoration of CFTR function in virtually all affected organs. When researchers in the mid-1990 s started to set up high-throughput assays to mitigate the basic defect in CF, they pursued the Garrodian approach to develop therapies for a specific class of mutations or just the major mutation F508del. Theratyping of non-F508del mutations and the uniform treatment of all CFTR genotypes by the expanded French Compassionate Program [76, 217–229, 250, 251] have now taught us that ELX/TEZ/IVA will improve the function of virtually all CFTR mutants if CFTR protein is synthesized in sufficient amounts. The researchers at Vertex have screened more than a million compounds in their high-throughput assays, 12 molecules of which were later examined in clinical trials with pwCF [8]. This highly selected set of molecules apparently recognizes the critical Achilles heels of proper processing, trafficking and function of the CFTR protein. Recombinant ELX/TEZ/IVA-bound F508del CFTR showed wild-type conformations in cryo-EM [146]. The binding sites for elexacaftor, tezacaftor and ivacaftor are not in close proximity to the position of phenylalanine 508 in NBD1. Thus, ELX/TEZ/IVA induces not only in F508del CFTR, but also in many more missense mutants close-to-normal conformations. Guided by the Garrodian approach of mutation-specific therapy, Oslerian-type correctors and potentiators emerged from the evolutionary race of small molecules in the high-throughput assays that fit (almost) all mutants. EMA’s recent approval of ELX/TEZ/IVA for all pwCF who produce mutant CFTR protein rewards the efforts of the CF community of the last five years to demonstrate the efficacy of triple therapy for non-F508del mutations.

On the other hand, we observe substantial heterogeneity of the responses of the individual patient in CFTR biomarkers and clinical outcomes (cf. Figures 1, 2). The biomarkers demonstrate normalization of CFTR activity on the population level, but fail to predict the association between CFTR activity and clinical outcome. The next years will tell us whether early treatment of children with minimal pathology of the affected organs will generate more coherent presentations of basic defect and of clinical phenotype. During pre-modulator times, CF disease showed personalized signatures that were more shaped by genetic modifiers, infection, inflammation, socioeconomic status and therapeutic intervention than by the disease-causing CFTR mutations [65, 66]. The authors are curious which types of personal signatures of CF disease will emerge during permanent treatment with efficient CFTR modulators. The Garrodian-type CFTR biomarkers of the next generation should address the long-term course of airway and gut microbiome [196, 211, 308], host defense, epithelial homeostasis [73] and multiorgan metabolism [117].

Abbreviations

AAV

Adeno-associated virus

ABPA

Allergic bronchopulmonary aspergillosis

ALI

Airway-liquid interface

ASL

Airway surface liquid

ASO

Antisense oligonucleotide

BIH

Berlin Institute of Health

CAVD

Congenital absence of the vas deferens

CF

Cystic fibrosis

CFF

Cystic Fibrosis Foundation

CFTR

Cystic fibrosis transmembrane conductance regulator

CFTR-RD

CFTR-related disorder

DIDS

4,4'-Diisothiocyanostilbene-2, 2'-disulfonic acid

DVA

Deutivacaftor

DZL

German Center of Lung Research

ELX

Elexacaftor

EM

Electron microscopy

EMA

European Medicines Agency

ER

Endoplasmic reticulum

FDA

US Food and Drug Administration

FIS

Forskolin-induced swelling

FRT

Fischer Rat Thyroid

HGVS

Human genome variant sequence

IBMX

Isobutylmethylxanthin

ICM

Intestinal current measurements

ICL4

Intracellular loop 4

IL

Interleukin

IQR

Intraquartile range

IVA

Ivacaftor

LUM

Lumacaftor

MAPK

Mitogen-activated protein kinase

MSD

Membrane-spanning domain

NBD

Nucleotide-binding domain

NPD

Nasal transepithelial potential difference

PI

Exocrine pancreatic insufficiency

PS

Exocrine pancreatic sufficiency

PTC

Premature termination codon

pwCF

People with cystic fibrosis

QPIT

Quantitative pilocarpine iontophoresis sweat test

RD

Regulatory domain

ROMA

Rectal organoid morphology analysis

SCC

Sweat chloride concentration

scRNA-seq

Single cell RNA sequencing

SERCA

SarcoEndoplasmic Reticulum Calcium ATPase

TEZ

Tezacaftor

TM

Transmembrane segment

TNF-α

Tumor necrosis factor

VAN

Vanzacaftor

VUS

Variant of unclear clinical significance

Authors’ contributions

BT designed the outline of the review, performed the literature search and wrote the first draft. BT, STP, AMD, SG, LN, OS, MAM acquired, analyzed and interpreted the data that were generated during the CFTR biomarker trials on CFTR modulators. All authors substantively revised the text, approved the submitted version and agreed to be personally accountable for their contributions to the work.

Funding

Open Access funding enabled and organized by Projekt DEAL. Authors’ published work on the impact of CFTR modulators on CFTR biomarkers has been supported by grants of the Bundesministerium für Bildung und Forschung (BMBF) to the German Center of Lung Research (DZL) (82DZL009B1, 82DZL002 A1, 82DZL005B1, 82DZL004B1), an independent medical grant from Vertex Pharmaceuticals Incorporated (IIS- 2018–107555) and the German Research Foundation (DFG) (CRC 900, A2 and Z1, CRC 1449, Z2).

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

B.T. reports grants and contracts from the German Research Foundation (DFG), the German Federal Ministry of Education and Research (BMBF) and the Volkswagen Stiftung with payments made to the institution; personal fees for advisory board participation and educational events from Vertex Pharmaceuticals; lecture honoraria from Vertex Pharmaceuticals and streamedup! GmbH.

S.T.P. reports grants and contracts from the Mukoviszidose e.V., the Else Kröner Fresenius Foundation and the Deutsche Forschungsgemeinschaft (DFG) with payments made to the institution and travel support from Vertex Pharmaceuticals. A-M.D. reports grants and contracts from the German Federal Ministry of Education and Research (BMBF), Vertex Pharmaceuticals Incorporated, European Cystic Fibrosis Society Clinical Trial Network (ECFS-CTN) and Christiane Herzog Stiftung with payments made to the institution, personal fees from the c4c consortium, GSK and European Cystic Fibrosis Society.

S.Y.G. reports grants from Mukoviszidose e.V. (German CF Foundation) and Vertex Pharmaceuticals Incorporated with payments made to institution; personal fees for advisory board participation from Chiesi GmbH and Vertex Pharmaceuticals Incorporated; lecture honoraria and honoraria for a CME module from Vertex Pharmaceuticals Incorporated.

L.N. reports grants and contracts from the German Federal Ministry of Education and Research (BMBF), the European Cystic Fibrosis Society and Vertex Pharmaceuticals with payments made to the institution.

O. S. reports grants from the German Federal Ministry of Education and Research (BMBF) with payment made to the institution and honoraria from Vertex Pharmaceuticals Incorporated for lectures.

M.A.M. reports grants and contracts from the German Research Foundation (DFG), the German Federal Ministry of Education and Research (BMBF), Boehringer Ingelheim, Enterprise Therapeutics and Vertex Pharmaceuticals with payments made to the institution; personal fees for advisory board participation or consulting from Boehringer Ingelheim, Enterprise Therapeutics, Kither Biotech, Splisense, Vertex Pharmaceuticals; lecture honoraria from Vertex Pharmaceuticals; and travel support from Boehringer Ingelheim and Vertex Pharmaceuticals.