Gene Therapy for Cystic Fibrosis: Hurdles to Overcome for Successful Clinical Translation Gene Therapy for Cystic Fibrosis: Hurdles to Overcome for Successful Clinical Translation

Additional information is available at the end of the chapter Abstract Cystic fibrosis (CF) is a genetic disease that hampers the lung function. Despite that the main defective gene has been deeply characterized, some relevant concerns still need to be resolved before considering gene therapy as a realistic medical choice. One of the major issues that need to be strongly considered in order to succeed in the search for an effective gene therapy approach for CF is the design of the appropriate genetic material to be delivered. Other relevant factors to take into consideration include the design of safe and effective gene delivery systems, the biological barriers that need to be overcome in order to reach the nucleus of the target cells, and the problems related to the design of a drug formulation suitable for lung delivery purposes. Furthermore, some problems related to the commercialization of gene therapy products also need to be resolved. In this chapter, we discuss the up-to-date strategies to overcome such hurdles in order for gene therapy to become a routine treatment modality for CF.


Introduction
Cystic fibrosis (CF) is a rare disease with low prevalence caused by the dysfunction of the transmembrane conductance regulatory gene (CFTR). The most prevalent CFTR mutation consists of a deletion of a phenylalanine at position 508 [1]. The disease presents a heterogeneous distribution in the world population being more frequent in Northern Europe.
According to recent reports, in the European Union, 1 of every 2000-3000 newborns is affected by CF. In the USA, the incidence is 1 per 3500 births. In Asia, the predisposition to CF is low; however, there is evidence to show that this disease is severely underdiagnosed [2]. The basic characteristic of CF is the transport of defective ions in the apical membrane of most secretory cells, which leads to an altered secretion of mucus in the epithelium of the respiratory tract, the digestive tract, the pancreas, the liver, and the reproductive track [1]. The conventional treatments available on market, which include, among others, antibiotics, pancreatic enzyme supplements, high-fat diets, and even physiotherapy [3], afford the consequences derived from CFTR dysfunction and have significantly improved the mean life expectancy of patients affected by the disease up to 34 years [4]. However, their quality of life is severely compromised mainly due to side effects and interactions among such treatments [5]. Therefore, other therapeutic options such as gene therapy, in which the main goal is to restore the function of the mutated CFTR protein acting on the genetic cause of the problem, need to be considered. CFTR gene was cloned more than two decades ago, and the monogenic and autosomal recessive nature of CF disease means that the addition and expression of the corrected gene could reverse the underlying cause of the disease. Therefore, there is reasonable hope to consider gene therapy as a potential realistic medical option, and consequently, some clinical trials have been performed since 1993. However, despite the moderate optimism that emerged with the development of such clinical assays, there are still some hurdles to overcome before considering gene therapy a realistic medical option. Main concerns are related to the intrinsic properties of genetic materials, the development of safe and efficient gene delivery vectors able to deliver genetic materials to the nucleus of target cells, the design of a drug formulation suitable for pulmonary gene delivery applications, and the hurdles associated with the commercialization of such drugs (Figure 1). Figure 1. Hurdles that gene therapy should overcome in order to reach clinical practice in the treatment of CF disease.
In the next sections, we will analyze such barriers along with the most relevant approaches developed by the scientific community to circumvent them in order to cure CF with gene therapy.

Plasmid DNA
Bacterial plasmid DNA (pDNA) remains an interesting biomolecule for gene transfer, with several promising reports and clinical trials in progress worldwide [6]. In CF, pDNA has been successfully delivered by nonviral vectors to the sheep lung [7]. Additionally, when administered in multiple-dosage regimen, no loss of activity was observed [8]. In order to be produced in recombinant bacteria and to express their therapeutic gene of interest (GOI), pDNA needs a bacterial origin of replication sequence (bac-ORI). In addition, pDNA backbone includes a sequence with resistance to one/various antibiotics such as kanamycin, which allows to select the clone of bacteria transformed that expresses the plasmid. Finally, a eukaryotic promoter is needed to enhance GOI expression [9]. Usually, when pDNA reaches the nucleus of target cells, it remains in an episomal position, which means that it replicates independently from the host chromosomal DNA, avoiding the undesirable activation of oncogenic genes [10]. The main concerns of pDNA in gene therapy are related to safety issues. Classically, in eukaryotic cells, pDNA has been associated with the induction of undesired immune responses and secretion of proinflammatory cytokines [11]. For instance, a transient neutrophilic infiltration and an elevation in proinflammatory cytokines have been reported in mouse lung [12]. Although the episomal nature of pDNA could be an interesting advantage, the transfection efficiency remains compromised mainly by the transient and relatively low gene expression. Additionally, the size of the plasmid, determined by the number of base pairs, jeopardizes transfection efficiency [13,14].

Minicircle DNA
In order to overcome the previously mentioned disadvantages associated with the use of pDNA in gene therapy, small plasmidic cassettes known as minicircle DNAs (mcDNAs) have been recently developed [15]. Cameron and Scheleff first employed mcDNA terminology in 1995. Nowadays, this technology offers a potential alternative to enhance both transfection efficiency and safety of gene delivery [14]. Basically, mcDNAs are circular constructors similar to pDNA but significantly smaller, since mcDNAs contain a minimal expression cassette, of a promoter, a transgene, and a polyadenylation, signal but are devoid of bacterial pDNA elements. Thus, mcDNA technology allows sustained transgene expression mainly due to a lower activation of nuclear transgene silencing mechanisms and reduced immunogenic responses in vivo [16,17].
In the lung, some promising results have been obtained with the use of small plasmidic cassettes [18]. In fact, results of a Phase IIb double-blind clinical trial for CF have been recently reported. These trials were performed with a plasmid encoding CFTR gene and lacking CpG bacterial region, known as pGM169 [8]. In such study, treated patients exhibited modest but significant improvements in lung function compared to placebo-treated ones during 1-year follow-up [19]. In any case, despite the optimism generated, there are still some concerns that need to be considered, such as the reproducibility of the results; the intensity of the response, probably conditioned by the degradation of formulation after aerosolization process; or the number of patients that received such treatment.

Genome editing tools
Both previously mentioned approaches based on pDNA and mcDNA technologies allow to restore the function of the mutated CFTR gene, with the addition of normal copies, but they do not correct the mutation at their local chromosomal location. However, genome editing tools based on zinc-finger nucleases (ZFNs), or transcription activator-like effector nucleases (TALENs), can specifically correct CFTR gene mutations at their natural chromosomal location, and so, the corrected gene can remain under the control of its endogenous promoter [20].
ZFNs are synthetic restriction enzymes, which have three or more zinc-finger DNA-binding motifs linked to the FokI restriction enzyme that recognizes trinucleotides in a specific DNA sequence [21]. When FokI enzyme creates a double-strand break (DSB) near the mutation place, cellular DNA repair mechanisms are activated to maintain cell viability. In these conditions, a donor DNA sequence with high 5′ and 3′ homology with the DNA sequence where DSB has been generated can be exogenously supplemented to enhance the correction of the mutation by homologous recombination (HR) mechanism. This genome editing tool has been successfully used in vitro to correct CFTRΔF508 mutation in both human bronchial epithelial cells [22] and CF-induced pluripotent stem (IPS) cells [23].
TALEN technology is very similar to ZFNs. These nucleases were originally characterized in Xanthomonas bacteria, in which TALEN proteins are secreted when Xanthomonas infect a wide variety of plants, thus activating genes that help to develop the pathogenesis. This genome editing tool also produces a DSB around the mutation site of the target gene and consequently induces cellular DNA repair mechanisms [24]. TALENs are considered as a more efficient and cost-effective alternative to ZFNs [25]. In the case of ZFNs, each finger module recognizes three to four bases of the DNA sequence. However, in the case of TALENs, gene recognition is mediated by a more specific mechanism, where each module of 33-35 amino acid targets a single nucleotide. This technology has been recently applied to correct CFTRΔF508 mutations in CF patient-specific IPS cells [26]. Overall, such study reported correction of patient-specific IPS cells in less than 3 months, which could allow rapid scaling up for future applications.
Clustered regularly interspaced short palindromic repeats (CRISPR) methodology, originally described as an adaptive immune response in archaea, follows the same rationale described for ZFNs and TALENs, but instead of protein domains, short RNA molecules are used to drive the required homology [27]. In this case, an endonuclease called Cas9 is guided by a single guide RNA (gRNA) to hybridize specifically with the mutated sequence in the DNA; then, as described for ZFNs and TALENs, the resulting DSB triggers cellular DNA repair mechanism [28]. The main advantage of CRISPR technology is that it is an easy-to-synthesize costeffective tool that is able to correct more than one mutation at the same time, if multiple-gene targeted sgRNAs are delivered to target cell along with the Cas9 protein, which makes it an excellent option [24]. CRISPR technology has been applied to repair CFTRΔF508 mutations in intestinal stem cell organoids of CF patients [29].
This study represents an interesting proof of concept for CFTRΔF508 correction by HR using CRISPR/Cas9 technology in primary adult stem cells derived from patients with a singlegene hereditary defect and offers reasonable hope to be successfully applied to the lungs of patients affected by CF. However, some relevant concerns, mainly related to the frequency of undesirable off targets, still need to be resolved in order to reach clinical practice [19].

Vectors
One of the main concerns related to the clinical application of gene therapy is the design and development of safe and effective gene delivery vectors to introduce exogenous genetic material into the nucleus of target cells [30,31]. In the absence of gene delivery vectors, naked genetic material is quickly degraded mainly by exogenous deoxyribonuclease enzymes, which clearly inhibit transfection efficiency [31]. Additionally, the negatively charged genetic material, mainly due to the phosphate groups, hampers the electrostatic interactions with cell membranes, which are negatively charged too. Therefore, the clinical application of gene therapy demands the design, characterization, and evaluation of efficient and safe carriers to mammalian cells.

Viral vectors
At present, viral-based carriers are the most appropriate from an effectiveness point of view. The natural evolution that viruses have undergone over many years has allowed them to face different intra-and extracellular barriers and, consequently, infect target cells with high efficiency.
In the CF field, a wide variety of viral-based vectors has been developed in clinical trials. The first one was performed in 1993 with adenovirus in three patients, where partial correction of the chloride transport in nasal epithelium was observed [32]. Some of the main advantages of adenoviruses include their non integrating nature and their natural tropism for the lung. However, despite such favorable properties, and the high transduction efficiency observed in most tissues, gene expression usually remains transient, and these viruses can induce strong immune and inflammatory responses in a dose-dependent manner, which clearly brings up safety issues and, therefore, limits their application in the clinical practice [8].
Initial clinical trials performed with adenovirus allowed the development of adeno-associated viruses (AAV), which have interesting characteristics for their application in gene therapy, such as broad tissue tropism, high transduction efficieny, and persistent episomal expression, which can last for years, even though it is a non integrating vector [33,34]. In addition, recombinant AAV vectors have been shown to be safe in several clinical trials, as they are not related to any known human disease. However, these vectors also present relevant limitations, the main one being their low capacity to load genetic material (<5 kb) [35]. Between 1999 and 2007, six clinical trials were conducted with these kinds of vectors in CF [8]. Phase I clinical trials demonstrated that a single-dose administration of AAV in the respiratory tract of patients affected by CF was safe and well tolerated [36]. Nevertheless, subsequent studies, with repeated doses in more patients, did not report significant improvement in lung function [37]. This lack of efficacy was mainly attributed to the low DNA loading capacity of AAV, which prevented loading the 4.7 kb of the CFTR gene [8]. In addition, AAV capsid-specific immune responses limited repeated administrations in patients [8]. Nowadays, some interesting strategies are being developed in order to minimize adaptive immune responses after repeated administration, such as the design of hybrid AAV capsids or the removal of CpG bacterial regions from AAV vectors [38,39].
Lentiviruses have an integrative nature and have shown long-term and stable transgene expression when administered in the respiratory tract of mice, which minimizes the need for repeated administration [40]. Additionally, the packaging of full-length CFTR gene and promoters is not limited by size. Therefore, nowadays they are considered promising vectors for the treatment of CF [41]. However, in order to consider its use in clinical practice, some concerns still need to be resolved, such as the scaling in the production of these vectors and the control of the place where the transgene is inserted into the genome of the pulmonary cells, which could increase the tumorigenicity potential of such viral vectors due to random integration [20]. Consequently, such viral vectors could be more suitable for ex vivo than for in vivo therapy. In any case, a promising study in three newborn CF pigs has recently shown that 2 weeks after lentiviral delivery by aerosolization, the anion channel defect can be corrected in a large animal CF model [42]. Other recent studies assessed with pseudotyped lentivirus vectors in both murine lungs and human air-liquid interface cultures showed that preexisting and acquired immune responses do not interfere with vector efficacy [43]. In such study, at least 14% of the airway cells were transduced. Interestingly, toxicological results, notably the integration site profile showing absence of integration near oncogenic loci, support further progression toward clinical trials.

Nonviral vectors
Although the use of viral-based vectors in clinical trials still predominates over that of nonviral vectors, in recent years, there has been a notable increase in preclinical studies using nonviral vectors [44]. The reason is that these systems represent a safer, cheaper, and easier to produce alternative to viral-based vectors [18]. The main advantages of nonviral vectors include, among others, the ability to produce them on a large scale with high reproducibility and low cost; their relative stability after storage; the possibility of multiple-dose regimen administration due to their low immunogenicity; their high capacity to carry genetic material, independently of the size [45]; as well as the possibility to modify them chemically in order to regulate important physicochemical parameters, such as size, charge, morphology, or polydispersion, which clearly influence their final biological properties. All these important advantages have raised the interest of the scientific community to develop new biocompatible materials of different structures, compositions, sizes, and characteristics to transport therapeutic genes into specific organs or cells, overcoming the different extra-and intracellular barriers [46].
Within the large variety of nonviral vectors developed, most of them are based on peptides as well as on cationic lipids and polymers, which form the corresponding complexes (polyplexes and lipoplexes) after electrostatic binding with DNA [47]. The resulting complexes protect nucleic acids from enzymatic degradation and facilitate cellular uptake by interactions with the cytoplasmic membrane [48]. The PEG-CK30 peptide, due to its low immunogenicity and its ability to be endocytosed by cells, is one of the most widely used, although the formulation must be optimized to allow its administration in aerosol form to reach the lungs [49].
Regarding cationic polymers, polyethylenimine (PEI) is one of the most used, since its chemical structure can be easily modified to increase the efficiency of transfection, for example, by incorporating lactose (Lac-PEI) to improve intracellular trafficking [50]. However, PEI has not yet been used in any clinical trial. The main limitation lies in the difficulty that exists to prepare PEI polyplexes at high DNA concentrations [49]. One of the most promising strategies that have been used to circumvent this problem is the use of ultrafiltration methods, through which PEI/DNA concentrates are prepared.
In the case of cationic lipids, some of the most widely used to develop nonviral formulations in the CF field are N-[1-(2,3-dioleyloxy) propyl]-N, N, N-trimethylammonium, dioleoylphosphatidyl ethanolamine, and dioleoyl trimethyl ammonium [51]. However, currently, the most promising nonviral vector in CF clinical trials is based on the cationic lipid GL6TA, which was synthesized to prevent DNA degradation in the lysosome and to be stable after pulmonary administration by aerosolization [8,51]. In 2011, a preclinical comparative study of the use of PEI, PEG-CK30, or GL67A nonviral vectors in aerosols demonstrated that this last formulation was the best one to transport DNA to sheep's lungs [7]. Moreover, in an extensive preclinical study performed in 2014 [52], it was corroborated that the formulation based on the lipid GL67A, which had already been used for 15 years in CF clinical trials, was still suitable for administration in multiple-dose regimen, without any observed loss of activity [8].

Biological barriers
In order to reach the nucleus of target cells and initiate transgene expression, the genetic material must overcome some extracellular and intracellular barriers, which will be discussed in this section, along with the most relevant strategies that have been developed to make the transfection process more efficient.

Extracellular barriers
Even though intravenous injection is one of the most commonly used administration routes, especially for delivering genetic cargo into cancerous cells, some barriers still hamper its use in clinical practice, particularly in the CF disease [53]. First of all, the genetic material needs to be protected against extracellular enzymatic digestion, since DNA is quickly degraded when administered alone. To avoid such quick degradation, nonviral vectors based on both positively charged lipids and polymers offer the possibility to condense on their surface the genetic material by electrostatic interactions and minimize such undesirable effect [54]. However, the final positive charge of polyplexes or lipoplexes can interact in a nonspecific way, not only with target cell membranes but also with other negatively charged components such as serum proteins that jeopardize transfection efficiency [55]. These interactions could result in the formation of aggregates that classically are eliminated from the blood by the reticuloendothelial system. Interestingly, the incorporation of polyethylene glycol (PEG) motifs into the formulation of some nonviral carriers enhances the stability of complexes, since the highly hydrophobic nature of PEG chains creates a steric barrier to prevent both aggregation of complexes in blood circulation and extracellular enzymatic degradation by nucleases [55]. In any case, other relevant aspects related with both the length and the degree of PEGylation should be also considered, as they can also decrease DNA condensation efficiency with nanoparticles [56]. In addition to PEG, other polymers with hydrophobic nature such as poly(4-acryloylmorpholine) or poly(N,N-dimethylacrylamide) have recently emerged as interesting and promising alternatives to compensate or ameliorate the negative effects associated with PEGylation [57].
Since intravenous injections present relevant hurdles that hamper the delivery of genetic material into target lung cells, local administration into the lung seems to be a reasonable alternative. In this case, the presence of mucus and the clearance mechanism are the most relevant barriers to overcome [58]. To avoid such barriers associated with pulmonary administration, other interesting noninvasive routes of administration, such as intranasal instillation, can be used to target lung cells. Nevertheless, the main problem is the low amount of genetic material that can be administered by the intranasal route. From a technical point of view, aerosolized nonviral vector/DNA complexes, carefully designed for inhalation in combination with appropriate excipients to enhance both particle flow and aerodynamic diameter, could be an interesting option since they are needle-free systems able to deliver locally high cargo concentrations [53].
To circumvent the diffusion of complexes into lung cells due to unspecific interactions with the biopolymer network of the mucus, some mucolytic agents that hydrolyze mucins can be added [58]. Other strategies include the incorporation of N-acetylcysteine to reduce disulfide bridges between the subunits of mucin, and consequently the viscosity [59], or the functionalization of nonviral vector formulations with mucolytic agents.

Intracellular barriers
Once extracellular barriers are overcome, there is still a long way full of hurdles before reaching the nucleus of target cells. Firstly, complexes carrying the genetic material need to be endocytosed by target cells. The interaction between complexes and cell membranes can occur in an unspecific way or can be mediated by a specific ligand, which is the preferred one, especially for in vivo applications [60]. Of note, the choice of ligand to be incorporated into the nanoparticle formulation depends not only on the target cell but also on the type of cell entry pathway that will be used once the ligand binds to the desired receptor.
Classically, there are four main pathways of endocytosis: clathrin-mediated endocytosis (CME), caveolae-mediated endocytosis (CVME), phagocytosis, and macropinocytosis [61,62]. CME is an energy-dependent mechanism widely studied and characterized [63]. Typically, this pathway is directly associated with lysosomes, where the genetic material needs to leave such biological compartment quickly before being degraded by the acidic environment and the enzymes found in lysosomes [63]. To avoid this enzymatic degradation, some interesting strategies can be used, such as the formation of pores in the endosome membrane by incorporation of amphiphilic cationic peptides. This creates strong internal tensions in the membrane and enhances the exit of endosome content through such pores [64]. Another strategy is known as the "proton sponge effect," where the low pH within the endosome allows the protonation of trapped compounds that have a large capacity to absorb protons from the medium (buffer effect). Such protonation causes an important entry of ions (H + and Cl − ) and water in the endosome, which produce a swelling effect and its rupture. This effect has been observed in some cationic polymers with high buffering capacity over a wide pH range [64]. In CVME, internalized molecules go to the caveosome instead of lysosome [61] avoiding lysosomal degradation; however, there is still ongoing debate, with some authors claiming that CVME can fuse with lysosomes [65].
Phagocytosis is a special type of endocytosis used mainly by macrophages, monocytes, neutrophils, and dendritic cells, although other cell types can also use this cellular entry pathway [61]. Endocytosis mediated by phagocytosis comprises the formation of membrane extensions with certain forms to capture particles generally greater than 1 μm. In contrast, for endocytosis mediated by macropinocytosis, membrane extensions do not surround particles but form some kind of protuberances that finally fuse with the cytoplasmic membrane. In many cases, the physicochemical properties of nonviral vector-based nanoparticles, such as particle size, superficial charge, morphology, or polydispersity, directly influence the endocytosis mechanism and consequently the transfection efficiency.
Once the DNA is released into the cytosol of cells, it must enter the nucleus to produce its effect. This is considered a significant barrier that nonviral vectors must overcome in order to mediate a good transfection efficiency. One commonly used strategy to enhance nuclear import of genetic material is to incorporate a nuclear localization signal (NLS), such as polylysine or protamine [66]. NLS contains some amino acids that interact with some proteins of the cytoplasm known as importines. These importins enhance nuclear entry through the nuclear pore complex of the nuclear membrane through an energy-dependent mechanism [67].

Drug formulation
The airways seem to be the natural way to treat respiratory diseases and a good alternative to systemic and more invasive procedures. Currently, aerosolization is the prefered method of administration for airway targeting since it is a noninvasive route that induces little stress to patients. Moreover, high quantities of drug can be deposited directly and fast into the lungs, which circumvents the blood circulation and avoids the first-pass effect of the liver. However, the effectiveness of such approach strongly depends on the development of smart drug formulation strategies. One of the critical steps that need to be taken into account for a successful gene delivery approach by inhalation is the formulation of the drug molecules into an appropriate inhalable form with sufficient stability and adequate aerodynamic properties [68]. Highly susceptible molecules, such as nucleic acid, require special attention when delivered by this route of administration. The physicochemical constraints such as the hydrodynamic shear forces generated during aerosolization can induce degradation of the nucleic acids, which will be more or less important depending on their size [69,70]. Therefore, the need to develop a suitable formulation able to protect the material from degradation and at the same time ensure delivery of nucleic acid to the target cells in the lung needs to be deeply considered. In this sense, the commonly accepted aerodynamic size for pulmonary gene delivery is within the range of 1-5 μm. Larger particles (4-7 μm) tend to deposit in the airways, while smaller particles (1-3 μm) and those in submicron range (<1 μm) reach the lower airways and deeper lung [68]. The aerodynamic diameter of a particle can be modified not only by changing its size but also by varying its density or shape, which opens new possible strategies for gene delivery to the lung, such as the design of large porous hollow particles [71].
Suitable formulations for pulmonary delivery are mainly prepared either by dissolving or by suspending the therapeutic molecules in a liquid or formulating them into a dry powder for inhalation using liquid inhalers (including nebulizers), dry powder inhalers (DPIs), or pressurized metered dose inhalers (pMDIs); each of them is suitable for different applications. Once the aerosolized droplets or microparticles are deposited next to the target location into the lungs, they need to dissolve in the lung lining fluid for subsequent absorption and cellular uptake [72]. Nowadays, viral gene delivery to the lungs is limited to liquid formulations using a nebulizer [36], and there is no dry powder or metered dose inhaler formulation available for any vector-drug combination. In most cases, the gene transfer efficiency to lung cells using viral vectors is still too low with traditional nebulizer devices, probably due to the degradation of viral envelope by the shear forces caused during aerosolization [73] and the viscous mucus found in obstructive diseases, like CF [74]. Moreover, and as previously highlighted, the use of adenoviral or AAV vectors would likely induce an acute immune response upon the initial administration or result in low efficacy following repeat dosing. This is particularly relevant since aerosolized gene therapy might require repeat dosing because mucus clearance mechanisms and/or phagocytes may engulf and destroy the drug vector before it can be taken up by target cells [75]. In contrast, the simpler composition of nonviral vectors may have, in this case, an advantage over viral vectors, making readministration potentially more successful.
Although pulmonary gene-based therapies have not yet been granted marketing approval, numerous strategies are being tested both in vitro and in vivo, and various clinical trials are underway [19,36]. Table 1 summarizes some of the strategies used to date for the pulmonary delivery of nucleic acids by aerosolization.
Nowadays, the most studied approach for gene delivery to the lung involves the nebulization of the selected formulation [76], turning it from a liquid solution to microdroplets. Depending on the aerosolization system used, such as jet, ultrasonic, or mesh nebulizers, the implemented hydrodynamic stress that the therapeutic molecules would be subject to varies [77]. Interestingly, several strategies have been studied to reduce the damage to the genetic material during the aerosolization process, by condensing the nucleic acids with positively charged molecules, such as polyethylenimine (PEI), protamine, or poly-L-lysine (PLL), among others [78].

Commercialization
In addition to the above concerns, other relevant issues specifically related to the commercialization of gene therapy medicinal products (GTMP) must also be considered. Commercially available medical products based on gene therapy along with cell therapy and tissue engineering are classified as advanced therapy medicinal products. Although highly promising, their translation into clinical practice is nowadays hampered by major critical issues such as complex regulatory and ethical aspects, along with the intrinsic difficulties to scale up these products to an industrial level [20].
Regarding the regulatory concerns of GTMP that affect clinical applications, the economical investments, along with their manufacture and control, demand more attention than chemically synthesized small molecules [86]. Therefore, a deep analysis of both costs and benefits needs to be done before considering the commercialization of such therapies [87].
Another relevant concern that jeopardizes the clinical use of GTMP in CF is the ethical aspect of clinical trials. Since the early 1990s, more than 25 Phase I gene therapy clinical trials have been conducted. These trials have been carried out largely to assess the safety and feasibility of gene transfer methods and their expression in the host, reporting variable successes for both viral and nonviral approaches. Gene therapy products designed for the treatment of CF must meet certain requirements in order to become a viable therapeutic option. For instance, their clinical efficacy must be demonstrated by analyzing appropriate variables of the lung function such as the patient´s vital capacity that they are able to expire in the first second of forced expiration (FEV1), their age, sex or body composition, and the therapeutic efficacy which must be maintained with repeated administrations. In addition, the GTMP must demonstrate an acceptable profile when it comes to side effects, and other considerations such as treatment of early versus established lung disease must also be analysed.
Since a high percentage of patients affected by CF are children, clinical trials involving these patients must carefully balance the potential benefits of these therapies and the associated risks [88]. Regarding this controversial issue, the Gene Therapy Advisory Committee recommends that clinical trials on children should only be performed under specific circumstances, whereby: (i) it has been demonstrated that the research is necessary to promote the health of the trial population, (ii) the research cannot be done in adults, and (iii) there is a high potential of therapeutic benefit [88]. In fact, owing to a demonstrated benefit of early gene therapy intervention, the age of enrolment of children in clinical trials has progressively reduced over the years from 18 to 12 years old. However, parents should have legal rights to make the final decision on behalf of their children.
Another critical hurdle that strongly compromises the clinical application of gene therapy products for the treatment of CF is the difficulty to scale up formulations that were originally developed for basic clinical research [89]. Most of these products are usually developed by small-and medium-sized enterprises, in collaboration with academic groups, which are usually highly engaged in preclinical activities, but have limited manufacturing experience at industrial level. For instance, the normal procedure for preparing nonviral-based gene therapy products is by simply mixing and pipetting the negatively charged genetic material and the positively charged polymer -or lipid-based nonviral vector formulations, which are often produced in the laboratory at small volumes that usually oscillate between 1 and 5 mL. However, the standardization of this procedure at industrial level to produce high and stable levels of complexes under GMP conditions represents a great challenge that needs to be overcome for successful clinical application. In this sense, pilot plants, which employ small volumes of the product, represent an interesting option to gain knowledge on the technical process before full scale up production.

Conclusion
Despite the fact that the CFTR gene was cloned two decades ago, the current, conventional treatments for CF focus on masking the main symptoms, rather than addressing the underlying genetic cause of the disease. In this sense, gene therapy represents a promising alternative to tackle CF, considering the autosomal recessive nature of the most relevant ΔF508 mutation. Although the main objective of gene therapy seems simple, there are some hurdles that need to be overcome before gene therapy for CF becomes a realistic treatment option. In any case, the increase in knowledge and recent advances in biopharmaceutical technology offer reasonable hope for the treatment of this devastating disease. The minicircle technology, along with the new gene editing tools, offer important advantages compared with classical plasmids used to add functional copies of the gene. Additionally, intense research in novel nonviral vectors functionalized to overcome both extra-and intracellular barriers and the possibility to aerosolize such formulations without losing activity merit special attention.