Treating Cystic Fibrosis with Gene Therapy

Treating Cystic Fibrosis with Gene Therapy In this dissertation we shall consider the field of gene therapy in specific relation to cystic fibrosis.   We examine the different delivery vector mechanisms that have already been explored and concentrate primarily on the adeno-associated vectors. We examine the current state of research and consider the advantages and drawbacks of the various methods considered. We conclude with a discussion and analysis of our findings and   make anumber of assumptions relating to the future direction of the researchin the field.   The rate of progress in the field of gene therapy has been enormous. We must remind ourselves that the first clinical gene transfer experiment took place in 1989 when a patient with malignant melanoma received genetically modified auto logous T cells. (Rosenberg SA et al1990) Gene therapy encompasses two major areas. The in vivo field, where genes are incorporated into the target cells of the living body and the ex-vivo field where the target cells themselves are genetically modified outside the body and then re-implanted. Medical science has been using the basic techniques of gene transfer for a long time. The technique has been exploited when viral genes are introduced to human cells when a viral vaccine is administered. The key technologies that allowed the transition from vaccination to gene therapy were the evolution of methods that allowed the genes to be isolated and replicated (cloned) and manipulated (engineered) prior to transfer into human cells (Freeman SM et al 1996) The key principle in this process is the efficient transfer of the manipulated therapeutic genes into the nuclei of target cells usually be means of various vectors. This dissertation will be considering the utilisation of these vectors in some detail. In broad terms, the new or modified genetic material is able to produce new proteins which can restore deficient or abnormal functions of genetically diseased tissues, to generate tissues that have entirely new properties or to create transplantable tissues for the controlled release of therapeutic proteins. (Russell SJ  1997) In terms of viral vectors, prior to 1996 science was dependent on the use of modified retroviral vectors (eg.MMLV) to effect gene transfer into the chromosomes of a target cell and the adenovirus vectors when such integration was not needed. Neither vector was particularly successful as the intact nuclear membrane (in then on-dividing state) was a major barrier for chromosomal gene integration. (Sikorski R et al 1998). A breakthrough came with the realisation that lentiviruses (e.g. HIV) have the same ability to transfer genetic coding into the cellular genome but could do this in the non-dividing or dormant phase cells. (Amado R G et al 1999) In vitro, lentiviruses have been shown to change the target cell’s expression of proteins for up to six months. importantly, they can be used for terminally differentiated cells such as respiratory epithelium. The only cells that the lentivirus cannot penetrate the nucleus are those in the quiescent (G0) state as this blocks the reverse transcription stages of protein synthesis. (Amado R G et al1999) Cystic fibrosis Cystic fibrosis is the most commonly lethal inherited recessive inthe caucasian population. It affects about 1 per 2,500 livebirths. Thetreatment of cystic fibrosis has improved enormously in the last fiftyyears with the life expectancy increasing from an average 10 years to30-40 years now. The prime cause of death in affected individuals is the repeatedcycle of infection, inflammation and fibrosis of the respiratory tractwhich eventually culminates in respiratory failure and death. The disease itself is caused   by mutations in the single gene forthe cystic fibrosis transmembrane conductance regulator (CFTR) whichproduces a protein found in sweat and pancreatic ducts, gut,seminiferous tubules, lungs and many other tissues. The mutationsresult in an abnormal protein which, when expressed in the lungs,produces thick viscous and dehydrated secretions. This does not allow for the efficient expulsion of bacterial pathogensfrom the lungs and a number of highly resistant forms of bacteria arecommonly found in cystic fibrosis (viz pseudomonas aeruginosa)(Porteous DJ et al 1997). An individual must receive a defective copy of the cystic fibrosisgene from each parent in order to develop the clinical picture ofcystic fibrosis. Following normal genetic principles, if two carriersconceive a child, there is a 25% chance that it will have cysticfibrosis, a 50% chance that it will be a carrier and a 25% chance thatit will have two normal cystic fibrosis genes. Viral and non viral vectors Viruses have an ability to enter a host cell and combine their owngenetic material with that of the host cell. This is the basicrationale behind the science of gene transfer therapy.   As we shalldiscuss in some detail in this dissertation, there are a number ofpotential viral vectors that have been explored, evaluated andexploited in the search for an efficient and safe form of therapy.Viruses are not the only vector that can be utilised . Simply placingDNA in the nasal mucosa will produce some incorporation into theepithelial cells (Knowles MR et al 1998). This â€Å"absorption† can bedemonstrably enhanced further by the combination of the DNA withvarious plasmid or lipid complexes(Zabner et al 1997) The advantages of lipid or plasmid assisted transfer mechanisms arethat they do not appear to generate the immunological responses thatare seen with the viral vectors. They can also be used to facilitatethe transport of much larger pieces of DNA which would otherwise belimited by the packaging consideration incumbent on the viral vectors.(Felgner P 1997). The use of retroviral vectors is far from straightforward. The heavilypublicised case in April 2000 brought some of the problems to theattention of the media. A retroviral manipulation of   a case of X-SCID(X linked severe combined immunodeficiency) was treated by gene therapywith an apparent degree of success (BBC 2002). This particular diseaseprocess is caused by a mutation on the gene which codes for the C chainof the cytokine receptors which is situated on the X chromosome andvital for the functional development of T Killer lymphocytes which aretherefore completely absent in the condition A multinational team used a retroviral vector to insert a functionalcopy of the gene into bone marrow stem cells which were thenre-transfused back into the patient. (Cavazzana-Calvo M et al 2000).This particular case resulted in a return to normal levels of T cellsin a comparatively short period of time. This was hailed in both thepopular media and the peer reviewed journals as a major success and itcan indeed be considered a landmark as it pioneered the successful useof an ex-vivo procedure that avoided direct in vivo transfer of thevector. The reason for specifically highlighting this particular case isthat following the initial optimism of the clinical team, two of thefirst ten patients with this condition who were treated in the same waysubsequently developed a leukaemia-like illness. Genetic analysis ofthe malignant cells suggested that the retroviral vector used in thetransfer had also activated an oncogene LIM-only2 (LMO2) which is knownto be associated with some forms of leukaemia. The clinicians reviewingthe situation felt that, although it was not the only cause of themalignancy it was one of the events that triggered it. Similar concernshave been raised in respect of other clinical trials. (Lehrman S 1999) The prime reason for presenting these events is to demonstrate thefact that there is both a theoretical and practical risk of insertionalmutagenesis. Reduction of the risk requires greater specificity of thetargeting of the genetic deficit   perhaps by directing the expressionof the therapeutic genes to various specific tissues utilising bothtransductional and transcriptional targeting. Relph K et al 2004), In terms of specific considerations of the arguments in favour of theuse of retroviral   vectors, one can cite the fact that they have ahighly efficient mechanism of gene transfer together with lowimmunogenicity. It is a well researched and well studied system and isknown to selectively infect actively dividing cells. The conversearguments reflect their disadvantages including their ability todisturb or activate oncogenes, the fact that they are difficulty tospecifically target and it is difficult to obtain high titres in theclinical situation (after Olsen, J. C. 1998). In broad terms, the principles behind the use of retroviral vectorsare that they must be modified in order not to be able to transmit anyovertly pathological coding. This involves the deletion of viral helpergenes such as gag, pol and env   to render the replication processinvalid. This is done by utilisation of a producer or packaging cellline. (Nichols, E. K 1998). An example of a commonly utilised and extensively researched vector isthe MoMuLV. It is an engineered vector which can store 8 kb of RNAwithout compromising packaging efficiency. It is a hybrid cell lineeasily grown in mouse fibroblast cells There is a subdivision of the retroviral vectors known as thelentivirus, which is the only retroviral vector capable of integratinginto the chromosomes of non-dividing cells. This has been effectivelydemonstrated in vitro (Naldini L et al 1996).    The biggest problem with the lentivirus vectors is that theyappeared to only produce very low titres. Some recent researchsuggested that a modification to a amphotropic envelope protien wascapable of allowing higher titre levels. (Rolls M et al 1999) At about the same time that the scientific press was learning aboutthe problems with retroviral transfer (see above) other investigatorswere working with adeno-associated viruses (AAVs). A similar processwas invoked using adeno-associated viruses to correct a genetic defectinvolving coagulation factor IX. The adeno-associated viruses were usedas they were considered to be amongst the safest candidates for genetransfer. They do not naturally cause disease processes in humans andhave only rarely been found to incorporate in a random fashion into thehuman genome. Although it is noted that adenoviruses do cause oncogeneactivation in rodents although it has not been found in humans(Blacklow NR 1988).   The trial had a very positive outcome. (Kay MA et al 2000), but thetrial author (in later research work) published a study which suggestedthat, in study mice, the vector used in the trials actually integrateditself into gene containing regions of DNA more frequently that it didinto non-coding regions (Kay MA et al 2003). The findings were reportedas the fact that new genetic material was randomly distributed amongstall of the chromosomes particularly at sites of gene activity. On thisbasis, there appears to be at least a theoretical basis for thepossibility of similar cellular defects such as occurred in the X-SCIDpatients. Adenoviruses are comparatively simple structures. They arecategorised as double stranded DNA viruses. They have icosahedralcapsids with twelve vertices and seven surface proteins. The virionitself is spherical and non-enveloped and in the region of 70-90 nm insize. Their natural history is that they are spread easily in the naturalstate by the faeco-oral route and also by respiratory inhalation whichclearly has great implications for the treatment of cystic fibrosis. A theoretical analysis would immediately suggest that the adenovirus should be a suitable candidate for gene therapy as they can codefor specific proteins and they do not produce infection pathogenicviral offspring.    The early trials into this particular area were reviewed by Griesenbach(Griesenbach U et al 2002) who pointed out that the cystic fibrosisgene was first cloned in 1989 and in the subsequent   years, 18different trials were carried out, all with rather low degrees ofsuccess. They collectively trialed three different vectors, namelyadenoviruses, adeno-associated viruses 2 and cationic liposomes, andalmost universally found that each vector had a very low rate ofclinically significant gene transfer and none was sufficient to achieveclinical benefit    Plasmid Complexes At its most basic level, a plasmid is a small accessory collection ofDNA which is found in the cytoplasm outside of the nucleus. They arecapable of independent replication and can be manipulated with rathermore ease than nuclear DNA. Early investigations into the field of gene transfer explored thepossibility of plasmid vectors   and demonstrated the feasibility of themethod to effect CFTR gene transfer in vitro (Alton EW 1993). Otherteams had demonstrated the fact that, in clinical use theplasmid-liposome is both nontoxic and non-immunogenic (Hyde,SC et al1993).   This appeared to raise the possibility that many of theimmunological problems encountered by teams working with viral mediatedgene transfer mechanisms might be circumvented. In vivo work (Yoshimura,K et al. 1992) had demonstrated that genescould be transferred into the cytoplasm by this method and Stribling, R(et al 1992) demonstrated that, once there, they would then replicatenormally. Alton experimented with a CFTR-plasmid preparation in miceand demonstrated that it was capable of correcting the chloride levelsin cystic fibrosis mice back to normal levels (Alton EW 1993) Although the initial results were encouraging, clinical trials weredisappointing as the plasmid complex could not easily penetrate thethick mucous residues in the diseased lungs of patients with cysticfibrosis. (Erickson,R 1993) The plasmids typically have a positively charged head-group which isable to bind to the DNA strand and a hydrophobic tail group whichfacilitates the transfer of the complex across the cellular membranes.Initial studies suggest that   between 100-1000 times more DNA isrequired to effect successful gene transfer when this method iscompared to viral vectors. (Santis,G et al 1994). One alternative adaptation has been reported by Stern M (et al 2003)who points out that one of the solution of delivery is to ensure thatthe respiratory epithelium is exposed to the DNA over a long period.Their solution was to encapsulate the CFTR-plasmid in a slow releasebiocompatible polymer. Clinical trials are underway but not yetreported. The adeno-associated vectors appear to have (at least on atheoretical basis) a number of advantages over the vectors that we havealready discussed. They are based on a virus vector that is alreadynon-pathogenic (Berns, KI et al 1995) and has a mechanism that allowsit to be a long-term persistent entity in human cells (Blacklow, NR etal 1989). The adeno-associated vectors are particularly useful indealing with disease process that involve single gene mutations. This,therefore   makes it particularly suitable for single gene disorderssuch as cystic fibrosis and alpha 1 antitrypsin deficiency. (Flotte, TRet al 1998). In addition, some workers have also developed vectors which are capableof   producing either inducible or constitutive expression of thecytokine, interlukin-10   (IL-10) which is an importantanti-inflammatory protein which, on a theoretical basis, could beuseful not only in cystic fibrosis   but in other disease process whichhave chronic inflammation as their prime manifestation (viz Type Idiabetes mellitus or inflammatory bowel disease) (Egan, M et al 1992).These manifestations have been studied and have now reached the stageof early clinical trials (Wagner J et al 2002). With specific reference to the implications of cystic fibrosis, wecan point to trials which have resulted in the expression of cysticfibrosis transmembrane conductance regulator (CFTR) from rAAV(recombinant adeno-associated vectors) in cell cultures (Flotte, TR etal 1993), in animal models (primates) (Afione, SA et al 1996), andagain in early phase I clinical trials (Wagner, J et al 1998) The rAAV-IL-10 model has been studied in bronchial cell culturesfrom cystic fibrosis patients, to determine the functional consequencesof CFTR complementation. This has not yet been demonstrated in vivowith humans but in both mice (Song, S et al 1998), and monkeys (Conrad,CK et al 1996) The overall results of these (and other) studies have shown that   itis possible to achieve long term gene transfer and functionalexpression of the replaced gene (some studies for as long as 18 months)without any overt pathological findings. The histological findings are something of a surprise however, as,at least in both primate and mouse studies, the vector-introduced DNAin this form does not appear to be assimilated into the geneticmaterial of the chromosome, but persists in log strings or concatemersthat are episomal, which is in complete contrast to what happens whenthe naturally occurring agent infects the cell. There is some evidenceto suggest that host cell intrinsic factors such as DNA-dependentprotein kinase play some role in this process (Song, S 2001). The significance of this finding could be that the exclusion of thefunctional, newly introduced DNA from the rest of the nuclear gene poolmay be less likely to produce effects that could be either potentiallydisruptive to the host cell and less likely to activate oncogenes.Phase I trials have demonstrated significant rises of CFTR levels inboth sinus and lung tissue with no evidence of vector-related toxicity.(Wagner, JA et al 1999)   The adeno-associated vectors are constructed from proviraladeno-associated vectors plasmids, which have the Rep and Cap proteinsdeleted and substituting the appropriate gene (CFTR or equivalent)between the rAAV2 inverted terminal repeats together with other signalsequences such as promoter and polyadenylation sequences (Flotte, TR etal 1994) The packaging processes allows for about 5 kb of rAAV genomes to becarried   in the vectors which are prepared using a cotransfectiontechnique utilising human embryonic kidney cells (HEK-293) where thevector plasmid is cotransfected into the cells with helper agents(plasmid pDG) being used to encode the rAAV2-rep and -cap genestogether with the adenovirus helper functions (Grimm, D et al 1998).These are incubated for between 48 and 72 hrs. The cells are then lysedand the resultant agents are then separated by ultracentrifugationagainst a density gradient and affinity chromatography (Zolotukhin, Set al 1999).   The vectors are thereby amenable to being separated by both theirphysical characteristics and also their biological characteristics(infectious units). They are carefully screened to ensure the absenceof any possible contamination from non-modified (replication competentAAVs) prior to clinical usage. (Muzyczka N 1994) The comparatively small â€Å"payload† of the adeno-associated vectorsis proving to be a significant problem. The vector itself is small whencompared to the comparatively large size of the CFTR gene. (Flotte TRet al 1993) It does not leave any room to manoeuvre to manipulate thevector-specific sequences in the way that we have described with theretroviral and adenoviral groups. (Flotte TR et al 2001). A number of authors have characterised the problem with theobservation that the rAAV is typically about 20 nm across which allowspackaging of about 4.7 kb (kilobases) of transferable modified gene(exogenous DNA). (Dong JY et al 1996), If   it is combined with otherenhancers such as the promoter, the polyadenylation signal, thisclearly reduces the capacity for the DNA component. (Duan D et al2000). The Yan paper (Yan Z et al 2000) has outlined a novelexploitation of the unique ability of the rAAV genomes to link togetherin strings which appears to have the ability to bypass this particularlimitation.( Flotte TR 2000). The mechanism itself is the capacity of two distinct rAAV genomes thathappen to simultaneously infect the same target cell to undergo anintermolecular recombination insider the transduced nucleus of thetarget cell. This was a chance finding which arose from work involvingrAAV-derived episomes (Kearns WG et al 1996) in primate airways. It wasfound that some of these episomes were configured as circular head totail concatemers (Duan D et al 1999). This could have been either froma â€Å"rolling circle† replication from a single vector or alternatively,from an intermolecular recombination of material from multiple cellularpenetrations which combined within the palindromic inverted terminalrepeat sequences that are an intrinsic part of the AAV genomestructure. The authors were of the opinion that it was likely to be thelatter eventuality (Duan D et al 1998) It was a logical progression to try to exploit this phenomenon andthereby bypass the limitations imposed by the relatively smallpackaging capacity of rAAV. The adeno-associated vectors capsid onlyhas a capacity of about 5 kb. If we consider that the 145 nucleotidestretch of the AAV-ITR (inverted terminal repeat) sequence has to be inplace at both ends of the single-strand DNA for the vector DNA to beboth replicated and packaged, this only leaves in the region of 4.7 kbof genetically active material in each rAAV particle. As we have cited earlier in relation to the Dong paper (Dong JY et al1996) the CFTR gene accounts for about 4.5 kb which leaves very littlespace for other enhancing material. Because of this, the actual CFTRvector that has been used in the clinical trials to date uses only theminimal promoter activity of the AAVs-ITR itself to actually activateand drive the CFTR expression (Flotte TR et al. 1993). To look at this potentially important development in a little moredetail we can consider Duan’s original paper (et al 2000) and theauthors describe what they call a â€Å"superenhancer†. They describe acombination of a potent simian virus (SV40) and CMV immediate earlyenhancer elements as being packaged in one rAAV vector and a luciferasegene assisted by a small minima;l promoter in another rAAV vector. Invitro experiments suggested that either the SV40 or the intrinsicpromoter activity of the AAV-ITR was sufficient for this purpose. Theintermolecular recombination described above, was found to occur inboth vitro and in vivo experiments and was found to be sufficient tohave a greater than additive effect. Initial results from these varying methods are encouraging insofaras they are producing results of transgene expression which are 100-600times greater than with the unenhanced vector alone. (Yan, Z et al2000) Although not directly referable to our considerations of cysticfibrosis, we should note that Yan’s group and other workers have doneexperimental work which has culminated in the long term expression offunctional levels of erythropoetin with this two vector method in micein vivo. (Naffakh N et al 1995), This basic principle has been further enhanced by Sun (Sun L et al2000) with an ingenious manipulation of the system. They triedinserting the promoter and the first half of the coding sequence in onerAAV vector, immediately followed by a splice donor and then theupstream half of an intron. In the other rAAV vector was the downstreamhalf of the intron, the splice acceptor, the second half of the geneand the polyadenylation signal. To quote the author verbatim: This strategy is efficient enough to mediate high-level expressionand the intermolecular junctions are apparently stable enough tomediate expression for several months in vivo. Although this is clearly an ingenious augmentation of the sameprinciple , we should note that there are both advantages anddisadvantages to both pathways. The strategy that adopts the superenhancer takes its strengths fromthe fact that the recombination mechanisms optimise theposition-independent and orientation-independent functions of theenhancers.   Consideration of the options would suggest that there arefour potential recombination outcomes from the process described.Either of the two vectors could be on the 5’ end of the heterodimericmolecule and clearly either molecule could be in either orientation. With the superenhancer option, all four of these possibleintermolecular recombination outcomes should be functional fortransgene expression whereas if compared to the split intron strategy,by using the same reasoning, it is clear that only one out of the fourcould work. On the other side of the argument, the superenhancer option has thedisadvantage that the actual coding sequence of the gene to betransferred must still fall within the packaging capacity of the vectoritself whereas the split intron allows for a greater functionalexpansion of the packaging capacity. (after Flotte TR et al 2000)   In either event it can be seen that these ingenious modificationseffectively eliminate the main size limitation of the rAAV deliverysystem. Although initial pre-clinical work is encouraging it appearsthat there is still some potential for a degree of immune responseparticularly if the host organism has not experienced the newlyproduced protein before. A number of studies have been done on animal (vertebrate andprimate) with only minimal success. Different administration methodshave been studied including direct administration into the lung (WagnerJ et al 1999), IM injection (Song, S et al 2001 B) and hepatic portalvein infusion (Song, S et al 2001 A)       Human clinical trials have taken place with these vectors (Flotte T etal 1996)(Wagner J et al 1998) (Virella-Lowell, I et al 2000). Thestudies were done on adult male and female patients (18-47 yrs) whowere pseudomonas free and had recently been hospitalised for IVantibiotic infusions The disappointing results were probably a reflection of the factthat the CFTR defect is also interconnected in some way with   aproinflammatory phenotype which appears to be triggered by the abnormalprotein via an unfolded protein response. The authors were able to showevidence that the rAAV-CFTR mechanism was able to   correct the proteinproduction defect, they found it clinically difficult to transduce asufficient number of cells in the airway to reverse the inflammatoryresponse. It is proposed to run further experimental work which combines the  CFTR expression with an anti inflammatory   gene such as the IL-10.There is some in vitro work to suggest that this may be a possibleworkable approach (Teramoto, S et al 1998). Other work on ways ofenhancing the phenotypic expression of the modified genotype hassuggested that the use of various promoters and the rAAV-CMV/beta-actinhybrid promoter (CB-AAT) was found to be tone of the most efficient, atleast when it was compared to the other tested options such as the CMV,E1, U1a and U1b promoter constructs (Teramoto, S et al 1998) Overall, the initial results appear to be encouraging. A singleinjection of an rAAV-CB-AAT vector in animal studies has resulted inhigh level, stable transgene expression which has persisted over thelife span of the experimental animals and that there was no detectableinflammatory response in the animals who had received this form oftreatment (Flotte TR 2002) Flotte (et al 2002) reports that four human clinical trials at bothPhase I and Phase II level are currently underway examining the effectsof the rAAV-CFTR vector. They had an entry cohort of seven patientswith the vector being applied to the nasal lining, the maxillary sinusand the bronchus. The authors report no adverse effects being found andthat they have observed transgene expression at doses of 6 x 108 drp inthe sinus or 1 x 1013 drp in the lung. There are no reported interimfindings from the Phase II trials as yet.    There is clearly a potential for clinical benefit on the basis of theresults found to date if one can extrapolate from in vitro and animalexperiments. The authors comment that, in contrast to the adenovirusvectors there is a marked lack of inflammatory toxicity with the rAAVvectors. Despite these positive comments, we should not, however, overlook thepotential limitations of this particular delivery system. These havebeen identified by various authors as:    The inhibitory effect of preexisting airway inflammation on rAAV transduction in the lungs (Virella-Lowell, I et al 2000) A relative paucity of receptors on the apical surface of airway epithelial cells (Summerford, C et al 1998), The relatively weak nature of the minimal promoters used in the first-generation rAAV-CFTR vectors(Flotte, TR et al 1993), The potential for adverse long-term effects from rAAV vector DNA persistence. (Wu, P et al 2000) The Flotte group are currently investigating this problem by examiningthe hypothesis that the barriers in the airways of the cystic fibrosissufferer are primarily   due to the neutrophil-derived -defensins (HNP1and HNP2) and are actually reversible by the mechanism of AAT proteindelivery (Virella-Lowell, I 2000) Wu and his co-workers have been looking at ways of manipulating thegenetic make up of the rAAV2 capsid and thereby trying to enhance thetargeting ability so that the vector specifically targets the serpinenzyme complex receptor on IB3–1 cells – which is virtually specificfor the Cystic fibrosis bronchial cells Zabner, J (et al 2000), have considered alternative rAAV serotypesin the hope of finding one that will bind more specifically to thebronchial cells Other peripheral adjuncts have also been explored includingpromoters to enhance the effects of complementation and superenhancerswhich have been shown to improve the ability of the rAAV toconcatermerise with the help of smaller amounts of promoter agents  (Duan, D et al 2000). Perhaps it is appropriate to conclude this section on considerationof adeno-associated vectors with a critical analysis of a very recentmulticentre, double-blind, placebo-controlled trial (Moss RB et al2004) This was a well constructed, fully statistically significant anddouble blinded trial which considered   both the safety and thetolerability of repeated doses of adeno-associated serotype 2 vectorrepeatedly given by aerosol inhalation. The vector contained â€Å"cysticfibrosis transmembrane conductance regulator (CFTR) complementary DNA(cDNA) [tgAAVCF], an adeno-associated virus (AAV) vector encoding thecomplete human CFTR cDNA.† The entry cohort was comparatively small with 42 patients, of whom20 received the active agent. A number of indices of airway functionwere measured. Of particular interest to our considerations in thisdissertation was the fact that   vector shedding was found in alltreated subjects up to 90 days after inoculation. And that all subjectswho received the active agent exhibited at least a fourfold increase inthe serum AAV2 neutralising antibody levels. Of the 20 treated patients, six subsequently underwent bronchoscopy.Of those six, gene transfer but not gene expression was demonstrated inall of them. On this basis, it would appear that the actual transfermechanism is effective, but there are other factors present whichappear to interfere with the subsequent expression of the gene in termsof protein production. The study did not comment on the possiblereasons for this. The authors were able to conclude that the delivery system workedwell with no evidence of adverse effects and that treated patientsdemonstrated an â€Å"encouraging trend in improvement in pulmonary functionin patients with CF and mild lung disease.† Lipid 67 We have discussed the various shortcomings of the virus-associatedvectors and this has prompted researchers to explore and consider otheroptimising options for facilitating gene transfer. Zabner (J et al1997) considered the use of cationic lipids in this process and foundone   GL-67:DOPE (colloquially known as lipid 67) which appeared to beparticularly helpful in the process. Cationic lipids appear to show a degree of promise as possible vectorsfor CFTR cDNA transfer into respiratory epithelial cells