Topical cystic fibrosis transmembrane conductance regulator gene replacement for cystic fibrosis-related lung disease

  • Review
  • Intervention

Authors

  • Tim WR Lee,

    Corresponding author
    1. A Floor, Clarendon Wing, Leeds General Infirmary, Leeds Regional Paediatric Cystic Fibrosis Centre, Leeds, West Yorkshire, UK
    • Tim WR Lee, Leeds Regional Paediatric Cystic Fibrosis Centre, A Floor, Clarendon Wing, Leeds General Infirmary, Great George Street, Leeds, West Yorkshire, LS1 3EX, UK. timlee@doctors.org.uk.

    Search for more papers by this author
  • Kevin W Southern

    1. University of Liverpool, Department of Women's and Children's Health, Liverpool, Merseyside, UK
    Search for more papers by this author

Abstract

Background

Cystic fibrosis is caused by a defective gene encoding a protein called the cystic fibrosis transmembrane conductance regulator (CFTR), and is characterised by chronic lung infection resulting in inflammation and progressive lung damage that results in a reduced life expectancy.

Objectives

To determine whether topical CFTR gene replacement therapy to the lungs in people with cystic fibrosis is associated with improvements in clinical outcomes, and to assess any adverse effects.

Search methods

We searched the Cochrane Cystic Fibrosis and Genetic Disorders Group Trials Register comprising references identified from comprehensive electronic database searches, handsearching relevant journals and abstract books of conference proceedings.

Date of most recent search: 22 August 2013.

An additional search of the National Institutes for Health (NIH) Genetic Modification Clinical Research Information System (GeMCRIS) was also performed for the years 1992 to 2013.

Date of most recent search: 04 September 2013.

Selection criteria

Randomised controlled trials comparing topical CFTR gene delivery to the lung, using either viral or non-viral delivery systems, with placebo or an alternative delivery system in people with confirmed cystic fibrosis.

Data collection and analysis

The authors independently extracted data and assessed study quality. Authors of included studies were contacted and asked for any available additional data. Meta-analysis was limited due to differing study designs.

Main results

Three randomised controlled trials met the inclusion criteria for this review, involving a total of 155 participants. Fourteen studies were excluded. The included studies differed in terms of CFTR gene replacement agent and study design, which limited the meta-analysis.

Although the first Moss study reported a significant improvement in respiratory function (forced expiratory volume at one second) 30 days after participants had received their first dose of gene therapy agent, this finding was not confirmed in their larger second study or in our meta-analysis.

In participants who received the CFTR gene transfer agents in the Alton study, "influenza-like" symptoms were found (risk ratio 7.00 (95% confidence interval 1.10 to 44.61)). There were no other significant increases in adverse events in any of the studies.

Alton measured ion transport in the lower airways and demonstrated significant changes toward normal values in the participants who received gene transfer agents (P < 0.0001), mean difference 6.86 (95% confidence interval 3.77 to 9.95). In these participants there was also evidence of increased salt transport in cells obtained by brushing the lower airway. These outcomes, whilst important, are not of direct clinical relevance.

Authors' conclusions

There is currently no evidence to support the use of CFTR gene transfer agents as a treatment for lung disease in people with cystic fibrosis. Future studies need to investigate clinically important outcome measures.

Résumé scientifique

Remplacement du gene régulateur de conductibilité transmembranaire pour la fibrose kystique pulmonaire.

Contexte

La fibrose kystique est provoquée par un gène défectueux codant une protéine appelée le régulateur de conductibilité transmembranaire de la fibrose kystique (RCTFK) et se caractérise par une infection pulmonaire chronique et des lésions pulmonaires évolutives qui entraînent une réduction de l'espérance de vie.

Objectifs

Déterminer si la thérapie de remplacement du gène RCTFK topique pour les poumons chez les personnes atteintes de fibrose kystique est associée à des améliorations de résultats cliniques et évaluer les effets indésirables éventuels.

Stratégie de recherche documentaire

Nous avons effectué une recherche dans le registre d’essais cliniques du groupe Cochrane sur la fibrose kystique et les troubles génétiques comprenant des références identifiées lors de recherches exhaustives dans des bases de données électroniques en effectuant des recherches manuelles de revues pertinentes et de résumés d’actes de conférence.

Date de la recherche la plus récente: 22 août 2013.

Une recherche supplémentaire du système d'information de recherches de mofification génétique (GeMCRIS Institut National de la Santé (NIH)) a également été effectuée pour les années 1992 à 2013.

Date de la recherche la plus récente: 4 septembre 2013.

Critères de sélection

Essais contrôlés randomisés comparant l'insertion topique de gènes RCTFK dans les poumons en utilisant des systèmes de transport viraux ou non viraux à un placebo ou à un autre système de transport chez des personnes atteintes d'une fibrose kystique confirmée.

Recueil et analyse des données

Les auteurs ont extrait les données et évalué la qualité des études de manière indépendante. Les auteurs des études incluses ont été contactés et les données supplémentaires éventuellement disponibles leur ont été demandées. La méta-analyse était limitée en raison des différentes conceptions des études.

Résultats principaux

Trois essais contrôlés randomisés, impliquant un total de 155 participants, ont rempli les critères d'inclusion de cette revue. Quatorze études ont été exclues. Les études incluses différaient concernant l'agent de remplacement du gène RCTFK ainsi que dans leur conceptions, ce qui limitait la méta-analyse.

Même si la première étude Moss a signalé une amélioration significative de la fonction respiratoire (volume expiratoire maximal par seconde) 30 jours après que les participants ont reçu leur première dose d'agent de thérapie génique, ce résultat n'a pas été confirmé dans leur deuxième étude qui était plus large ni dans notre méta-analyse.

Chez les participants ayant reçu les agents de transfert du gène RCTFK dans l'étude Alton, des symptômes «pseudo-grippaux» ont été observés (risque relatif 7,00 (intervalle de confiance à 95%, de 1,10 à 44,61)). Il n'y a pas eu d'autres augmentations significatives des événements indésirables dans aucune des études.

Alton a mesuré le transport d'ions dans les voies aériennes inférieures et a démontré des changements significatifs vers des valeurs normales chez les participants ayant reçu des agents de transfert de gènes (P < 0,0001), différence moyenne 6,86 (IC à 95%, de 3,77 à 9,95). Chez ces participants, des preuves d'une augmentation du transport de sels dans les cellules ont également été obtenues par brossage des voies aériennes inférieures. S' ils sont intéressants, ces résultats n'ont pas d'importance clinique directe.

Conclusions des auteurs

Il n'y a actuellement aucune preuve pour soutenir l'utilisation de réactifs d'agents de transfert de gènes comme traitement des maladies pulmonaires chez les personnes atteintes de fibrose kystique. Les études à venir doivent étudier les mesures de résultats cliniquement importants.

Plain language summary

Replacing the defective gene is a potential treatment for progressive lung disease in people with cystic fibrosis

In cystic fibrosis the gene encoding a protein called the cystic fibrosis transmembrane conductance regulator (CFTR) is faulty. People with cystic fibrosis suffer from progressive lung infection and damage which reduces life expectancy. Agents for CFTR gene transfer may be an effective treatment as they deliver a correct copy of the CFTR gene to lung cells.

We found three trials with 155 participants to include in this review. The trials compare gene therapy to placebo both of which are inhaled as a mist into the lungs. The trials were of different designs and used different agents. This meant we could not combine their data in a meta-analysis. We found significant changes toward normal values for the movement of molecules in the lower airways of people who received gene transfer agents. However, we found no evidence that outcome measures which are clinically relevant to people with cystic fibrosis had improved. In one study "influenza-like" symptoms were significantly more common in people who received CFTR gene transfer agents. We conclude that at the moment there is no evidence to support the use of CFTR gene transfer agents as a treatment for lung disease in people with cystic fibrosis. We recommend that future studies are designed and reported clearly so that their results can be incorporated into a systematic review.

Résumé simplifié

Le remplacement du gène défectueux est un traitement potentiel pour les maladies pulmonaires évolutives chez les personnes atteintes de fibrose kystique

Dans la fibrose kystique, le gène codant une protéine appelée le régulateur de conductibilité transmembranaire de la fibrose kystique (RCTFK) est défectueux. Les personnes atteintes de fibrose kystique souffrent d'infections pulmonaires progressives et des lésions qui réduisent l'espérance de vie. Les agents pour le transfert de gène RCTFK peuvent être un traitement efficace, car ils insèrent une copie correcte du gène RCTFK dans les cellules des poumons.

Nous avons identifié trois essais à inclure dans cette revue , avec 155 participants. Les essais comparent la thérapie génique à un placebo qui sont inhalés sous forme d'aérosol dans les poumons. Les essais avaient des conceptions différentes et utilisaient des agents différents. Cela signifie que nous n'avons pas pu les combiner dans une méta-analyse. Nous avons constaté des changements significatifs vers des valeurs normales pour le mouvement moléculaire dans les voies aériennes inférieures des personnes ayant reçu des agents de transfert de gènes. Cependant, nous n'avons trouvé aucune preuve indiquant qu'il y avait une amélioration de critères d'évaluation cliniquement pertinents pour les personnes atteintes de fibrose kystique. Dans une étude, les symptômes «pseudo-grippaux» ont été significativement plus fréquents chez les personnes ayant reçu des agents de transfert de gènes RCTFK. Nous concluons que, pour le moment, il n'y a aucune preuve pour soutenir l'utilisation d'agents de transfert de gènes RCTFK comme traitement des maladies pulmonaires chez les personnes atteintes de fibrose kystique. Nous recommandons que les études à venir soient conçues et notifiées de manière claire afin que leurs résultats puissent être intégrés dans une revue systématique.

Notes de traduction

Traduit par: French Cochrane Centre 14th January, 2014
Traduction financée par: Ministère du Travail, de l'Emploi et de la Santé Français

Background

Description of the condition

Cystic fibrosis (CF) is the commonest life-shortening disease in Caucasians and is caused by a single gene defect. It has a prevalence of approximately 1 in 2000 at birth (Bobadilla 2002). The affected gene is responsible for making a protein called the cystic fibrosis transmembrane conductance regulator (CFTR). There are many abnormalities or alterations of the CFTR gene that can affect the production and function of CFTR (Rowntree 2003). Normally, CFTR has an important role in co-ordinating salt transport across cell membranes. This role is particularly important in the lungs where an abnormality of CFTR results in dehydration of the surface liquid that lines the airways (Matsui 1998). Consequently the airway is not able to remove bacteria and chronic airway infection results in an intense local inflammation and mucus secretion (Boucher 2004). Subsequently a cycle of increasing lung inflammation and lung damage progressively reduces lung function, leading eventually to premature death (often in the third or fourth decade of life) from respiratory failure (FitzSimmons 1993; Frederiksen 1996).

Description of the intervention

These disease processes could be prevented by inserting a correct copy of the CFTR gene into the cells of people with CF, a process termed CFTR gene transfer therapy (Griesenbach 2004).

How the intervention might work

There are three main reasons why CFTR gene transfer therapy has been proposed for CF lung disease:

  1. the CFTR gene has been identified and characterised and it is possible to manufacture artificial CFTR genes (Riordan 1989);

  2. CF is a progressive condition with a potential window for early treatment when the lungs are relatively unaffected (Ranganathan 2004);

  3. delivery of gene therapy reagents to the lungs may be possible by aerosolisation or other methods of topical application (Lee 2005).

Early laboratory studies demonstrated the successful transfer of the CFTR gene to cells and animal airways using a variety of different gene delivery methods (Southern 1996). Essentially, delivery agents have been described as viral (where the CFTR gene is incorporated into a replication incompetent virus) and non-viral (most commonly positively-charged liposomes which when mixed with DNA increases uptake into cells) (Lee 2005). Human studies using viral vectors suggest that topical delivery to nasal airway cells results in uptake of the CFTR gene and expression of the gene by the airway cells (Knowles 1995; Zabner 1996). These studies demonstrate some correction of abnormal salt transport; however, repeat dosing results in inflammation (immune-mediated) and no detectable gene transfer (Harvey 1999; Zabner 1996). Repeated dosing of non-viral gene delivery agents does not result in such marked immune-mediated responses; however, steady-state levels of gene expression and functional changes are less impressive than those achievable acutely with viral vectors (Hyde 2000). Much effort is being directed towards developing gene transfer agents and strategies with both improved gene uptake and expression, and reduced immunogenicity (Ferrari 2002; Griesenbach 2004; Lee 2005).

Why it is important to do this review

This review is an updated version of previous Cochrane Reviews (Lee 2007; Lee 2012). It focuses on treatment of CF lung disease as this is the major cause of morbidity and mortality in people with CF, however abnormal CFTR function affects other parts of the body, some of which may be amenable to other forms of gene transfer therapy but will not be the subject of this review.

Objectives

To investigate whether topical CFTR gene replacement therapy to the lungs of people with CF is associated with improvements in clinical outcomes (respiratory function and quality of life) and assess existing or predicted adverse effects.

Methods

Criteria for considering studies for this review

Types of studies

Randomised controlled trials (RCTs), published or unpublished. Controlled clinical trials (CCTs), including quasi-randomised controlled trials, will be included in future updates if available and only if there is sufficient evidence that control and intervention groups were similar at baseline.

Types of participants

Children and adults with CF confirmed by the presence of two disease-causing mutations, or by a combination of positive sweat test and recognised clinical features of CF.

Types of interventions

Topical CFTR gene delivery to the lung, using either viral or non-viral vector systems, compared to placebo or an alternative vector system.

Types of outcome measures

Primary outcomes
  1. Respiratory exacerbations (assessed by need for additional oral or intravenous antibiotics)*

    1. oral antibiotics (days or episodes)

    2. intravenous antibiotics (days or episodes)

  2. Lung function testing (absolute values or percent predicted for age, sex and height)

    1. forced expiratory volume at one second (FEV1)

    2. forced vital capacity (FVC)

    3. relative change in FEV1 or FVC

    4. other relevant lung function tests (for example, infant lung function tests and lung clearance index)

  3. Survival (either as a binary outcome or time to event)*

Secondary outcomes
  1. Number of days as a hospital inpatient (or number of inpatient episodes)*

  2. Need for extra treatment*

    1. physiotherapy (duration or episodes)

  3. Adverse events*

    1. mild (e.g. sore throat, hoarse voice, dry mouth)

    2. moderate (e.g. wheeze, cough, fever, local allergic reaction)

    3. severe (e.g. coughing up blood (haemoptysis), collapsed lung (pneumothorax), chest infection, systemic allergic reactions)

  4. Quality of life (as measured by a validated and appropriate method)*

  5. School or work attendance*

  6. Nutritional parameters (including z scores or centiles)

    1. weight

    2. body mass index (BMI)

    3. height

  7. Acquisition of newly cultured respiratory pathogens

    1. Pseudomonas aeruginosa

    2. Staphylococcus aureus

    3. Haemophilus influenzae

  8. Sputum rheology (stickiness, as measured by a validated method)

  9. Mucociliary clearance of the airway (as measured by a validated method)

  10. Airway potential difference measurements (measuring salt transport)

    1. Baseline potential difference

    2. Response to perfusion with amiloride

    3. Response to perfusion with a zero chloride solution

  11. Measures of gene expression

    1. quantitative reverse transcription polymerase chain reaction (rtPCR) to measure vector-specific CFTR messenger ribonucleic acid (mRNA) production

    2. any other method of measuring gene expression

  12. Measures of CFTR protein expression

  13. Radiological measures of lung disease

    1. validated chest radiograph scores

    2. validated computerised tomogram (CT) score

*indicates outcomes that have a direct influence on the person with CF (and therefore are more pragmatic measures of efficacy).

Search methods for identification of studies

Electronic searches

Relevant studies were identified from the Group's Cystic Fibrosis Trials Register using the search term: topical (aerosolized) CFTR gene replacement for the lungs.

The Cystic Fibrosis Trials Register is compiled from electronic searches of the Cochrane Central Register of Controlled Trials (CENTRAL) (updated each new issue of The Cochrane Library), quarterly searches of MEDLINE, a search of Embase to 1995 and the prospective handsearching of two journals - Pediatric Pulmonology and the Journal of Cystic Fibrosis. Unpublished work is identified by searching the abstract books of three major cystic fibrosis conferences: the International Cystic Fibrosis Conference; the European Cystic Fibrosis Conference and the North American Cystic Fibrosis Conference. For full details of all searching activities for the register, please see the relevant sections of the Cystic Fibrosis and Genetic Disorders Group Module.

Date of latest search of the Group's Trials Register: 22 August 2013.

An additional search of the National Institutes for Health (NIH) Genetic Modification Clinical Research Information System (GeMCRIS) was also performed for the years 1992 to 2013 (Appendix 1).

Date of latest search: 4 September 2013.

Searching other resources

Investigators who work in the field and previous authors were also contacted for unpublished or additional data.

Data collection and analysis

Selection of studies

Two authors, TWRL and KWS, independently selected the studies to be included in the review. There was agreement between the authors on the suitability of studies for inclusion in the review.

Data extraction and management

The authors used a standardised form to extract data from studies eligible for inclusion in the review. If standard errors were reported, these were converted to standard deviations. Meta-analysis was undertaken on data from all included studies.

As the effect of gene transfer therapy may be cumulative, we analysed studies with different durations of intervention separately. For example, we planned to combine studies of less than one month duration, studies between one and two months duration, three and four months duration and five and six months duration. We planned to combine studies longer than six months duration on a three-monthly basis. Regarding studies of less than one month duration, in a post-hoc change to protocol, we did not consider it appropriate to combine six-hour data with Day 30 data, as data obtained within 24 hours of administration are related to monitoring of adverse effects rather than efficacy.

We report when measurements were taken by the primary investigators during the study, what measurements were reported within the published paper and what data we are reporting in the review (see Additional Tables (Table 1; Table 2)).

Table 1. Primary outcomes measured
  1. FEV1: forced expiratory volume in one second
    FVC: forced expiratory volume
    IV: intravenous

Study nameRespiratory exacerbationsLung functionDays in hospitalSurvival
Alton 1999Not measured.Measured, but only at 6 hours post-dose to monitor for adverse effects.

Reported in this review.
Not measured.Not measured.
Moss 2004Measured respiratory exacerbations requiring IV antibiotics within 150 days.

Reported in this review.
Measured FEV1, FEV1 % predicted, FVC, at baseline, 30, 60, 90 and 150 days.

Day 30 and 60 data reported in this review.
Not measured, although number of inpatient episodes within 150 days measured.

Reported in this review.
Not measured.
Moss 2007Measured respiratory exacerbations requiring IV antibiotics within 210 days.

Reported in this review.
Measured FEV1, FEV1 % predicted, FVC, at baseline, 14, 30, 45, 60, 75 and 90 days.

Day 30 and 60 data reported in this review.
Not measured.Not measured.
Table 2. Secondary outcomes measured
Study nameExtra treatmentAdverse eventsQuality of lifeSchool or work daysNutritionNew pathogensSputum rheologyMucus clearanceAirway PD
Alton 1999Not measured.Measured.

Reported in this review.
Not measured.Not measured.Not measured.Not measured.Not measured.Only measured saccharine nasal mucociliary clearance, this is not an efficacy measure for CFTR gene replacement therapy to the lung.

Not reported.
Measured baseline potential difference, response to perfusion with amiloride, and response to low chloride solution and isoprenaline, measured at baseline and 2 days after study drug.

Reported in review.
Moss 2004Not measured.Measured to 150 days.

Reported in this review.
Not measured.Not measured.Not measured.Measured Pseudomonas aeruginosa and Staphylococcus aureus, (at baseline and Day 90) but data is incomplete.

Reported in review.
Not measured.Not measured.Not measured.
Moss 2007Not measured.Measured to 210 days.

Reported in this review.
Not measured.Not measured.Not measured.Not measured.Not measured.Not measured.Not measured.

Assessment of risk of bias in included studies

In order to establish a risk of bias, two authors, TWRL and KWS, independently assessed the studies to be included in the review. The authors assessed the methodological quality of each study based on a method described by Jüni and colleagues (Jüni 2001). The authors assessed the following:

  • a description of the degree of blinding;

  • inclusion of all participants in an intention-to-treat analysis, regardless of whether they completed the treatment schedule or not;

  • whether there is a clear description of generation of allocation sequence, for example with a random numbers table;

  • whether concealment of allocation sequence is described, for example if neither investigators nor participants can foresee assignment to either treatment or control group.

There was agreement between the authors on the quality of studies for inclusion in the review.

Measures of treatment effect

For binary outcome measures, we calculated a pooled estimate of the treatment effect for each outcome across studies using the risk ratio. For continuous outcomes, we recorded either mean relative change from baseline for each group or mean post-treatment or intervention values and standard deviation. We calculated a pooled estimate of treatment effect by calculating the mean difference. For studies reporting continuous outcomes at multiple time points there are currently no methods to analyse aggregate longitudinal data if individual patient data are not available. Therefore in this review we have carried out analysis at each individual time point reported and assumed zero correlation between results. For any outcomes producing time-to-event data (e.g. survival) or count data (e.g. number of days), we analysed the data using the most appropriate method.

Unit of analysis issues

We planned to consider studies with a cross-over design for inclusion; however, we have concerns about the potential for 'hangover effects' following too brief a washout period (Southern 2003). This is particularly pertinent for gene transfer therapy agents which may alter the natural history of the CF airway disease. We would have analysed any data from cross-over studies that were eligible for inclusion using a method described by Elbourne (Elbourne 2002). Elbourne says that this approach will produce conservative results, as it does not take into account within-patient correlation. Also, each participant will appear in both the treatment and control group, so the two groups will not be independent. If cross-over studies are included in future updates of this review we will undertake a sensitivity analysis to determine the effect of these studies on the overall result.

Dealing with missing data

In order to allow an intention-to-treat analysis, we sought data on the number of participants with each outcome event, by allocated treated group, irrespective of compliance and whether or not the individual was later thought to be ineligible or otherwise excluded from treatment or follow-up.

Assessment of heterogeneity

We assessed the degree of statistical heterogeneity between studies using the I2 statistic, as described by Higgins, to establish if there are any inconsistencies between studies (Higgins 2003).

Assessment of reporting biases

In future updates, should sufficient studies become available, we will attempt to assess whether our review is subject to publication bias by using a funnel plot (which graphically illustrates variability between studies and should not demonstrate a systematic difference). If asymmetry is detected, causes other than publication bias will be explored.

Data synthesis

We have analysed the data included in the review using a fixed-effect model. If in future we identify a moderate or high degree of heterogeneity, we plan to analyse the data using a random-effects model.

Subgroup analysis and investigation of heterogeneity

If we had identified any heterogeneity, we would have investigated this using subgroup analysis of potential effect modifiers. The major potential effect modifier is age, other effects include sex and treatment centre.

Sensitivity analysis

If cross-over studies are included in future updates of this review we will undertake a sensitivity analysis to determine the effect of these studies on the overall result.

Results

Description of studies

Results of the search

Eighteen studies were identified in the searches, but only three were eligible for inclusion in the review. Fourteen studies were excluded from the results of the search. Further references identified indicated a multidose randomised controlled trial currently in progress, once this is completed we will assess its eligibility in a future update (Alton 2012).

Included studies

Three of the 18 studies identified in the searches were eligible for inclusion in the review (Alton 1999; Moss 2004; Moss 2007). The largest study has been published in full since the first version of this review appeared, but additional data had been previously provided by the research team (Moss 2007).

Trial Design

All three trials were randomised controlled trials of parallel design comparing CFTR gene replacement delivered to the lungs of people with CF to placebo (Alton 1999; Moss 2004; Moss 2007). The two Moss studies were multicentre, with the earlier one recruiting from eight centres and the later one from 12 centres (Moss 2004; Moss 2007); the Alton study was single centre (Alton 1999). The Moss studies were run in the USA (Moss 2004; Moss 2007) and the Alton trial was based in the UK (Alton 1999).

The later Moss study was the largest, recruiting 109 participants, although only 102 received at least one dose of the study drug (Moss 2007). The Alton trial was the smallest with only 16 participants (Alton 1999).

Both Moss trials report a power calculation where they calculated adequate power to test differences between treatment groups with respect to the primary and secondary protocol end points (Moss 2004; Moss 2007).

Follow up was shortest in the Alton trial at only 29 days (Alton 1999). Follow up in the 2004 Moss trial was longer at 150 days (Moss 2004). The third trial had varying length of follow up dependent on outcome; follow up was to 90 days for lung function and 210 days for adverse event monitoring (Moss 2007).

Participants

In total all three trials involved 155 participants with CF (Alton 1999; Moss 2004; Moss 2007). Alton recruited only males (Alton 1999). The two Moss trials recruited both males and females with approximately even gender split in each group, except for the treatment group in the 2004 Moss trial where there were more than double the number of females to males (14 females, 6 males) (Moss 2004; Moss 2007).

With regards to age, Alton specified that all participants in the trial were adults with a mean age of 26.9 years (Alton 1999). Both the Moss trials recruited participants of at least 12 years of age; the mean age of participants in the earlier trial was 23.7 years and in the later trial 22.6 years (Moss 2004; Moss 2007). A breakdown of the mean ages of participant by treatment group is available in the tables (Characteristics of included studies).

All trials reported the genotypes of participants. In the Alton trial, 12 out of 16 participants were homozygous ∆F508, but did not report the numbers for each mutation separately by treatment group (Alton 1999). Moss reported that in the earlier trial the placebo group had significantly more ΔF508 homozygous participants (77%) than the CFTR gene replacement group (25%) (P = 0.01) (Moss 2004). In the later trial, there were slightly more ΔF508 homozygous participants - 53% in each group (Moss 2007).

With regard to other participant characteristics, all trials stated a minimum FEV1 % predicted score; Alton stated this was at least 70% (Alton 1999) and for both Moss trials this was at least 60% (Moss 2004; Moss 2007). Alton also reported that there were no significant differences between groups in terms of body mass index, or chest X-ray score (Alton 1999). Both Moss trials also reported that both groups were similar in terms of demographics and clinical characteristics (Moss 2004; Moss 2007).

Interventions

All three trials compared an active treatment to a placebo (Alton 1999; Moss 2004; Moss 2007). Alton delivered the CFTR gene as plasmid DNA complexed with liposome (GL-67/DOPE/DMPE-PEG5000); in the treatment group 16 ml from a 20 ml solution containing 42.2 mg of plasmid DNA was nebulised once to the lungs, the placebo group received liposome alone (Alton 1999). Both Moss trials used the adeno-associated virus serotype 2 (AAV-2) vector tgAAVCF to deliver the CFTR gene (Moss 2004; Moss 2007). In the earlier trial, 1x1013 particles tgAAVCF or matching placebo were nebulised to the lungs on three occasions 30 days apart (Moss 2004). The later trial used a similar design, but used two instead of three doses, but still 30 days apart (Moss 2007).

Outcomes

Adverse events was the only outcome measured by all three trials (Alton 1999; Moss 2004; Moss 2007). Alton also reported on changes in measures of CFTR protein function (gene expression, CFTR protein expression and airway potential difference) (Alton 1999). The earlier Moss trial reported on clinical outcomes, such as change in lung function and high resolution computerised tomography (HRCT) lung scan scores, as well as respiratory exacerbations, inpatient episodes, acquisition of new pathogens and gene expression (Moss 2004). Moss also reported on respiratory exacerbations and lung function in the later trial (Moss 2007).

Excluded studies

Fourteen studies were excluded from the results of the search (see Characteristics of excluded studies). Eight of the excluded studies were not randomised controlled trials (Davies 2011; Flotte 1996; Harvey 1999; Joseph 2001; Noone 2000; Zabner 1996; Zabner 1997; Zuckerman 1999). The remaining six studies were excluded since the therapy was not applied directly to the lungs; four studies applied treatment to the nose (Gill 1997; Hyde 2000; Knowles 1995; Porteous 1997) and two studies applied treatment to the sinuses (Wagner 1999; Wagner 2002).

Ongoing studies

We have identified one ongoing trial (Alton 2012). This is a double-blind, randomised, placebo-controlled trial of parallel design. It is aiming to recruit 130 participants with confirmed CF aged 12 years and over. The active intervention is the aerosolised delivery of pGM169/GL67A to the lungs and this will be compared to placebo (0.9% saline). The dose schedule is 12 doses per patient, each of these four weeks apart over 48 weeks in total. The trial is planning to report on lung function, lung clearance index, HRCT score, respiratory exacerbations and safety. Results are expected late in 2014 or early 2015.

Risk of bias in included studies

Allocation

All three included studies were described as randomised, but none gave any specific details of the methodology (Alton 1999; Moss 2004; Moss 2007). None of the included studies gave any information regarding the concealment of allocation (Alton 1999; Moss 2004; Moss 2007). We therefore judged all three studies to have an unclear risk of bias as regards generation of allocation sequence and the concealment of allocation sequence.

Blinding

In all three studies there was double-blinding to the intervention (Alton 1999; Moss 2004; Moss 2007); however, only Alton explicitly stated that both participants and outcome assessors were blinded (Alton 1999). We deemed all three studies to have a low risk of bias from blinding.

Incomplete outcome data

In the Alton study, all randomised participants were assessed on an intention-to-treat basis, although data is incomplete for four of the reported endpoints (nasal histology, viability of lipid complex, chloride efflux and bacterial adherence) (Alton 1999). For these endpoints there is thus a potential risk of bias.

In the earlier Moss study, 42 participants were randomised, 37 (88%) received at least one dose of study drug and had outcome measures reported, and 35 (83%) completed all three doses of study agent (Moss 2004). It is not stated whether those who withdrew had been allocated the active or placebo treatment. Results were pooled for all 37 participants receiving at least one dose of study drug. Follow up was complete for these participants for most endpoints to 150 days, but only 35 participants underwent HRCT lung scans, and just 10 had vector-specific CFTR gene expression assessed, resulting in a high risk of potential bias (Moss 2004).

In the later Moss study, of the 109 participants randomised, 102 (94%) received at least one dose of study drug and had outcome measures reported (Moss 2007). Reasons for withdrawal or their treatment allocation were not reported for the seven participants who did not proceed to receive study drug post randomisation. Respiratory function data were reported to 90 days for all 102 participants receiving the study drug. Four participants receiving the study drug discontinued follow up after day 90 and before day 210; one in the active group was lost to follow up; and in the placebo group one died as a result of a motorcycle accident, one suffered an adverse event, and one was lost to follow up. Data on induced sputum inflammatory markers were reported for 97 participants receiving the study drug (89% of those randomised) (Moss 2007). We deemed this study to have a low risk of bias from incomplete outcome data.

Selective reporting

In the later Moss study, some relevant secondary endpoints, such as number of days of oral antibiotic use, remain unreported which suggests a potential risk of bias (Moss 2007).

Other potential sources of bias

In the earlier Moss study, the placebo group had significantly more ΔF508 homozygous participants (77%) than the CFTR gene replacement group (25%) (P = 0.01), which could affect the validity of this study due to the potential for gene replacement therapy to be more effective in participants with certain CFTR gene abnormalities versus others ( Moss 2004).

Effects of interventions

Primary outcomes

1. Respiratory exacerbations
a. Oral antibiotics (days or episodes)

This outcome was not reported in any of the included studies.

b. Intravenous antibiotics (days or episodes)

In the 2004 Moss study there was no reduction in respiratory exacerbations requiring intravenous antibiotics within 150 days, relative risk (RR) 1.70 (95% confidence interval (CI) 0.50 to 5.79) (n = 20 CFTR gene therapy and n = 17 placebo) (Moss 2004). Similar results were seen in the later Moss study within 210 days, RR 1.33 (95% CI 0.62 to 2.89) (n = 51 CFTR gene therapy and n = 51 placebo) (Moss 2007). As the studies examined different time scales (150 and 210 days) it is not possible to combine these data in a meta-analysis. The Alton study did not report this outcome measure (Alton 1999).

2. Lung function testing

Two studies reported lung function testing as a measure of effectiveness at time points between 14 and 150 days following initial dose (Moss 2004; Moss 2007).

The first study reported change in FEV1 and FVC from baseline at 30 days, 60 days, 90 days, and 150 days (Moss 2004). The second study reported change in FEV1 and FVC from baseline at 14 days, 30 days, 45 days, 60 days, 75 days and 90 days (Moss 2007). As the participants in the earlier Moss study, but not the later study, received a third dose of CFTR gene replacement or placebo on Day 60, meta-analysis was only possible on data from Day 30 and Day 60 compared to baseline, and not at subsequent time points.

One study reported lung function at six hours post dose to monitor for adverse effects (Alton 1999).

a. Forced expiratory volume at one second (FEV1)

Although the Alton study reported a fall in FEV1 at six hours post dose there was no significant difference between active and placebo groups, mean difference (MD) -1.4 (95% confidence intervals (CI) -3.07 to 0.27) (n = 8 CFTR gene therapy and n = 8 placebo) (Alton 1999).

In the later Moss study, there was no significant difference in FEV1 at Day 30 between the CFTR gene replacement group (mean (SD) 2.83 (0.76) L) versus the placebo group (2.93 (0.78) L, P = 0.50) (Moss 2007).

In the earlier Moss study, the change in absolute FEV1 at Day 30 from baseline was significantly improved, MD 0.14 (95% CI 0.02 to 0.26) (n = 20 CFTR gene therapy and n = 17 placebo) (Moss 2004). At the same time point, the later Moss study did not show a significant improvement in absolute FEV1 from baseline, MD 0.03 (95% CI -0.05 to 0.11) (n = 51 CFTR gene therapy and n = 51 placebo) (Moss 2007). Combined data for both Moss studies are presented and show no significant mean difference, MD 0.06 (95% CI 0.00 to 0.13) (n = 71 CFTR gene therapy and n = 68 placebo) (Analysis 1.2).

At Day 60, the earlier Moss study reported change in absolute FEV1 from baseline, the difference between the groups was not significant, MD 0.05 (95% CI -0.12 to 0.22) (n = 19 CFTR gene therapy and n = 17 placebo) (Moss 2004). At the same time point, the later Moss study also showed no significant difference, MD -0.07 (95% CI -0.15 to 0.01) (n = 51 CFTR gene therapy and n = 51 placebo) (Moss 2007). Combined data for both Moss studies at this time point are presented and show no significant mean difference, MD -0.05 (95% CI -0.12 to 0.02) (n = 70 CFTR gene therapy and n = 68 placebo) (Analysis 1.2).

At subsequent time points there were no significant differences in the change in absolute FEV1 from baseline demonstrated in either the earlier Moss study (90 days and 150 days) or the later Moss study (45 days, 60 days, 75 days and 90 days) (Moss 2004; Moss 2007); furthermore, these data cannot be pooled due to different study regimens.

The two Moss studies reported change in FEV1 % predicted as an outcome (Moss 2004; Moss 2007). The earlier Moss study reported change in FEV1 % predicted at 30 days, there was no significant difference between the groups, MD 2.99 (95% CI -0.44 to 6.42) (n = 20 CFTR gene therapy and n = 17 placebo) (Moss 2004). At the same time point the later Moss study also reported no significant difference between the groups, MD 0.80 (95% CI -1.50 to 3.10) (n = 51 CFTR gene therapy and n = 51 placebo) (Moss 2007). Combined data for the two Moss studies at this time point are presented and show no significant difference, MD 1.48 (95% CI -0.43 to 3.39) (n = 71 CFTR gene therapy and n = 68 placebo) (Analysis 1.3).

At 60 days the earlier Moss study reported change in FEV1 % predicted, with no significant MD 0.39 (95% CI -4.31 to 5.09) (n = 19 CFTR gene therapy and n = 17 placebo) (Moss 2004). At the same time point, the later Moss study significantly favoured placebo, MD -2.69 (95% CI -5.14 to -0.24) (n = 51 CFTR gene therapy and n = 51 placebo) (Moss 2007). Combined data for both Moss studies at this time point are presented and show no significant difference, MD -2.03% (95% CI -4.21 to 0.14) (n = 70 CFTR gene therapy and n = 68 placebo) (Analysis 1.3). At subsequent time points there were no significant differences in change in FEV1 % predicted from baseline demonstrated in either study and the data cannot be pooled for analysis (Moss 2004; Moss 2007).

b. Forced vital capacity (FVC)

Although the Alton study reported a fall in FVC at six hours post dose there was no significant difference between the groups, MD -1.70 (95% CI -3.27 to -0.13) (n = 8 CFTR gene therapy and n = 8 placebo) (Alton 1999).

Both Moss studies assessed the change in absolute value of FVC (L) from baseline (Moss 2004; Moss 2007). At Day 30, the earlier Moss study reported no significant difference between the groups, MD 0.13 (95% CI -0.02 to 0.28) (n = 20 CFTR gene therapy and n = 17 placebo) (Moss 2004). At this time point the later Moss study also demonstrated no significant difference between the groups MD -0.01 (95% CI -0.09 to 0.07) (n = 51 CFTR gene therapy and n = 51 placebo) (Moss 2007). Combined data for the two Moss studies at this time point are presented and show no significant difference, MD 0.02 (95% CI -0.05 to 0.09) (n = 71 CFTR gene therapy and n = 68 placebo) (Moss 2004; Moss 2007) (Analysis 1.4). At Day 60, the earlier Moss study reported no significant difference between the groups, MD 0.02 (95% CI -0.16 to 0.20) (n = 19 CFTR gene replacement and n = 17 placebo) (Moss 2004). The later Moss study also reported no significant difference between the groups, MD -0.07 (95% CI -0.15 to 0.01) (n = 51 CFTR gene therapy and n = 51 placebo) (Moss 2007). Data from these two studies at Day 60 are combined and show no significant difference, MD -0.06 (95% CI -0.13 to 0.02) (n = 70 CF gene and n = 68 placebo) (Analysis 1.4). At later time points there remained no significant difference in change in absolute value of FVC (L) from baseline when comparing CFTR gene replacement to placebo in either study, and these data were not suitable for pooling in Statistical Analysis.

c. Relative change in FEV1 or FVC

Individual participant data were not presented in any of the included studies and thus relative change and analysis using the generic inverse variance was not possible.

d. Other relevant lung function tests (for example, infant lung function tests and lung clearance index)

The Alton study reported a small but significant reduction in gas transfer in the CFTR gene replacement (mean (SD) 7.5% (7.6) but not the placebo group (1.6% (7.6), P < 0.05) when measured on Day 2. Alton reported this had returned to baseline values when measured on Day 29, but did not present any actual data (Alton 1999).

3. Survival*

Not reported as a specific outcome measure in any of the included studies. There were no deaths reported in either active or placebo groups in the Alton or the first Moss studies. In the later Moss study one participant in the placebo group died of an unrelated motorcycle accident (Moss 2007).

Secondary outcomes

1. Number of days as a hospital inpatient (or number of inpatient episodes)*

The number of days as a hospital inpatient was not reported by any of the studies. In the earlier Moss study there was no significant difference in number of inpatient episodes within 150 days, relative risk (RR) 0.85 (95% CI 0.37 to 1.94) (n = 20 CFTR gene therapy and n = 17 placebo) (Moss 2004). It is not stated clearly in the study report whether these episodes are independent from each other. This outcome measure was not reported by the other included studies (Alton 1999; Moss 2007).

2. Need for extra treatment*
a. Physiotherapy (duration or episodes)

This outcome was not reported in any of the included studies.

3. Adverse events*

All included studies reported adverse events as number of participants experiencing each adverse event, rather than absolute number of adverse event episodes. Due to the large number of different adverse events reported, we are only presenting significant results in this section. For further details, please see Analysis 1.6.

a. Mild (e.g. sore throat, hoarse voice, dry mouth)

In the Alton study there was no difference between the active and placebo groups in terms of mild adverse events, which were classed as cough; wheeze; or tight chest, all occurring within 48 hours of dosing (Alton 1999). Data were pooled, making comparisons with other studies difficult. In both Moss studies mild adverse events were individually recorded as rhinitis, pharyngitis, headache and sinusitis; again there were no significant differences between active and placebo groups reported (Moss 2004; Moss 2007).

b. Moderate (e.g. wheeze, cough, fever, local allergic reaction)

The Alton study reported a significant increase in influenza-type symptoms in the CFTR gene replacement group resolving within 30 hours of dosing, risk ratio (RR) 7.00 (95% CI 1.10 to 44.61). The earlier Moss study assessed abdominal pain, asthma, chest pain, cough, increased cough, dyspnoea, fatigue, fever, decreased lung function and increased sputum, with no significant differences between active and placebo groups in any outcome to 150 days (Moss 2004). The later Moss study assessed abdominal pain, asthma, increased cough, dyspnoea, fever, decreased lung function, and increased sputum (Moss 2007). This study also found no significant difference between groups to 210 days (Moss 2007).

c. Severe (e.g. coughing up blood (haemoptysis), collapsed lung (pneumothorax), chest infection, systemic allergic reactions)

The Alton group reported no severe adverse events in either CFTR gene replacement or placebo groups (Alton 1999). There were no significant differences in incidence of severe adverse events between the groups in either Moss study (Moss 2004; Moss 2007).

4. Quality of life*

This outcome was not reported in any of the included studies.

5. School or work attendance*

This outcome was not reported in any of the included studies.

6. Nutritional parameters
a. Weight

This outcome was not reported in any of the included studies.

b. Body mass index (BMI)

This outcome was not reported in any of the included studies.

c. Height

This outcome was not reported in any of the included studies.

7. Acquisition of newly cultured respiratory pathogens
a. Pseudomonas aeruginosa

The 2004 Moss study reported this endpoint, by culture of expectorated sputum at baseline and Day 90, but data are incomplete (Moss 2004). Of the 16 participants in the gene therapy group assessed at baseline, 11 were culture positive for Pseudomonas aeruginosa (P. aeruginosa) (68.8%). At Day 90, 14 participants from the gene therapy group were assessed and 10 of them were positive for P. aeruginosa (71.4%). In the placebo group, eight participants were assessed at baseline and four of these tested positive for P. aeruginosa (50%); at Day 90, twelve participants were assessed and eight of these tested positive for P. aeruginosa (66.7%). It is not possible to draw any conclusions from these limited data.

b. Staphylococcus aureus

Again, this endpoint was reported by the earlier Moss study, but data collection was incomplete (Moss 2004). Of the 16 participants in the gene therapy group assessed at baseline, six were culture positive for Staphylococcus aureus (37.5%). At Day 90, 14 participants from the gene therapy group were assessed and three of them were positive for Staphylococcus aureus (21.4%). In the placebo group, eight participants were assessed at baseline and three of these tested positive for Staphylococcus aureus (37.5%); at Day 90, of the twelve participants assessed seven tested positive for Staphylococcus aureus (58.3%). Again, it is not possible to draw any firm conclusions from these limited data.

c. Haemophilus influenzae

This outcome was not reported in any of the included studies.

8. Sputum rheology

This outcome was not reported in any of the included studies.

9. Mucociliary clearance of the airway

This outcome was not reported in any of the included studies. Alton reported simple saccharine nasal mucociliary clearance, but this is not an efficacy measure for CFTR gene replacement to the lung (Alton 1999).

10. Airway potential difference measurements (measuring salt transport)

One study reported airway potential difference measurements (Alton 1999). These were measured in the lower airways using a bronchoscope and catheter, once before the administration of the study drug, and again two days after the study drug was administered.

a. Baseline potential difference

People with CF have a higher baseline potential difference than people without CF (Griesenbach 2005). An effective CFTR gene replacement intervention would therefore be expected to demonstrate a reduction in baseline potential difference. No significant change in baseline potential difference was seen in either the CFTR gene replacement group (mean change -0.81 millivolt (mV), SD 9.02) or the placebo group (mean change 2.53 mV, SD 6.48), with no difference between the groups, MD -3.34 (95% CI -11.04 to 4.36) (n = 8 CFTR gene therapy and n = 8 placebo) (Alton 1999). Data presented here are for segmental bronchi, estimated from figures in the original paper (Alton 1999).

b. Response to perfusion with amiloride

Since sodium absorption is increased in people with CF, a greater reduction in potential difference from baseline is seen in response to amiloride in participants with CF than in those without CF (Griesenbach 2005). Thus effective CFTR gene replacement should reduce the fall in potential difference seen following amiloride. Again, there was no significant change in response to perfusion with amiloride in either the CFTR gene replacement group (mean change -12.6%, SD 14.66) or the placebo group (mean change -16.50%, SD 13.17), with no difference between the groups, MD 3.90 (95% CI -9.76 to 17.56) (n = 8 CFTR gene therapy and n = 8 placebo) (Alton 1999).

c. Response to perfusion with a zero chloride solution

The response to zero or low chloride secretion following prior perfusion with amiloride is the airway potential difference measurement that is the most effective discriminator between CF and non-CF participants (Alton 1999; Griesenbach 2005). Alton reported this as summed response to low chloride solution and isoprenaline. A separate control group of participants without CF had a mean response of 10.70 (SD 4.11). In the CFTR gene replacement group there was a significant improvement following perfusion with a zero chloride solution (mean 5.4 (SD 3.82) P < 0.05), with no significant change in the placebo group (mean -1.46 (SD 2.29). This change in the CFTR replacement group was significantly different from that in the placebo group (P < 0.0001), MD 6.86 (95% CI 3.77 to 9.95) (n = 8 CFTR gene therapy and n = 8 placebo). Data have been estimated from figures in the original paper (Alton 1999).

11. Measures of gene expression
a. Quantitative reverse transcription polymerase chain reaction (rtPCR) to measure vector-specific CFTR messenger ribonucleic acid (mRNA) production

This outcome was reported in two studies, but at different time points (Alton 1999; Moss 2004). In the Alton study bronchial brushings were performed two days following the single dose of CFTR gene replacement or placebo. No vector specific mRNA could be detected by this measure in either the CFTR gene replacement group (n = 8), or the placebo group (n = 8) (Alton 1999). In the Moss study six out of the twenty participants in the active group and two out of the seventeen participants in the placebo group had bronchial brushings performed 30 to 60 days after the third dose of study drug, again no vector specific mRNA could be detected in any participant sampled (Moss 2004).

b. Any other method of measuring gene expression

This outcome was not reported in any of the included studies.

12. Measures of CFTR protein expression

This outcome measure was not assessed in either of the Moss studies (Moss 2004; Moss 2007). Alton measured chloride efflux in cells obtained by bronchial brushing, by measuring changes in 6-methoxy-N-(-sulphopropyl) quinolium (SPQ) fluorescence (Alton 1999). Any increase in CFTR protein expression following the delivery of the study drug would result in an increased efflux of the SPQ fluorescence. There was no significant difference in SPQ efflux between the two groups, MD 0.3 mMs-1 (95% CI -0.12 to 0.72) (n = 6 CFTR gene therapy and n = 7 placebo). Data for the placebo group were estimated from a figure in the original paper. The original paper reports a significant increase in efflux (P < 0.05) in the active, but not in the placebo group (Alton 1999).

13. Radiological measures of lung disease
a. Validated chest radiograph scores

This outcome was only reported in the Alton study, who used the previously validated CF Northern score (Conway 1994). The investigators report "no significant change from baseline" at Day 29, but the data are not shown (Alton 1999).

b. Validated computerised tomogram (CT) score

Only the earlier Moss study reported this outcome measure (Moss 2004). Participants underwent a high resolution CT scan at baseline and on Day 90, 30 days after the third dose of study drug had been administered. The CT scans were scored on a scale from 0 to 100 using an algorithm that had not been previously validated and was developed for this study. There was no significant difference between the groups in terms of change in CT score from baseline at Day 90, with the CFTR gene replacement group mean (SD) -1.00 (3.00), placebo group mean (SD) 0.00 (6.00); MD -1.00 (95% CI -4.17 to 2.17) (n = 18 CFTR gene therapy and n = 17 placebo). These data must be treated with caution as the scoring system had not been previously validated.

Discussion

Three randomised controlled trials met the inclusion criteria for this review. They compared CFTR gene replacement to placebo, delivered to the lungs of participants with CF (Alton 1999; Moss 2004; Moss 2007). There were differences between the studies in terms of CFTR gene replacement agent and study design. Alton examined a single dose of non-viral liposome-based CFTR gene therapy agent with follow up for 29 days (Alton 1999). The Moss studies used a viral CFTR gene delivery system with follow up for 150 to 210 days respectively (Moss 2004; Moss 2007). These studies were of similar design, although three doses were given in the first study and two doses in the second. A number of distinct outcome measures of efficacy, generally of a surrogate nature, were employed by the two groups of investigators; however, the primary focus of the first two studies was safety.

All studies have a parallel design; however, the methods of randomisation and allocation concealment are unclear. In the earlier Moss study, only 83% of randomised participants received all three doses of study drug and it is not stated whether withdrawals had been originally allocated to the active or the placebo group (Moss 2004). Post-allocation withdrawal also occurred to a lesser extent in the later Moss study (Moss 2007). Given the small number of eligible studies, we have included these studies despite these issues with methodological quality; however, in future updates when more studies are included, we will examine the impact of these data on the meta-analysis with a sub-group analysis.

The most important result from these studies was the finding of "influenza-like" symptoms in participants who received the CFTR gene transfer reagents (seven out of eight participants) in the Alton study (Alton 1999). There were no other significant increases in adverse events in any of the studies. Alton measured ion transport in the lower airways and demonstrated significant changes toward normal values in the participants who received gene transfer reagents. In these participants they also demonstrated evidence of increased salt transport in cells obtained by brushing the lower airway in participants who had received gene transfer reagents compared to placebo.

The first Moss study demonstrated a significant improvement in respiratory function (FEV1) 30 days after participants had received their first dose of gene therapy agent; however, this finding was not confirmed in the larger second Moss study or in our meta-analysis (Moss 2004; Moss 2007). The Alton study assessed lung function at six hours post dosing as a marker of acute toxicity, rather than as a measure of efficacy which cannot be assessed until later time points (Alton 1999). Thus respiratory function data from the two Moss studies should not be compared to those from the Alton study.

Gene therapy has potential as a fundamental treatment modality for people with CF. To date nearly 400 people with cystic fibrosis have volunteered as participants for clinical studies of gene therapy. As studies evolve to examine gene transfer as a therapy for CF lung disease, it is imperative that these studies are designed and reported in a manner that is clear and enables them to be incorporated into a systematic review.

Authors' conclusions

Implications for practice

The evidence to date suggests that the topical administration of CFTR gene transfer agents has failed to restore the CFTR functional deficiency present in CF lung disease, and there is thus no current evidence to support the use of gene transfer reagents for this condition.

Implications for research

Clinical studies examining the use of CFTR gene transfer reagents for CF lung disease should include outcome measures that are pertinent to people with CF as well as clearly defined surrogate measures of efficacy. It is imperative that future studies of established or novel gene therapy reagents use a robust study design with clearly defined endpoints and adequate power. These studies must be reported in a full and transparent manner to enable the incorporation of data into systematic reviews.

Acknowledgements

We would like to acknowledge the Cochrane Cystic Fibrosis & Genetic Disorders Review Group, in particular Nikki Jahnke and Tracey Remmington. We thank Ashley Jones for invaluable statistical support and the investigators, Richard Moss and Eric Alton, and Alison Heald, Targeted Genetics Corporation, for their interest and help with the study.

Data and analyses

Download statistical data

Comparison 1. CFTR gene replacement therapy versus placebo
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Respiratory exacerbations (episodes)2 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
1.1 5 - 6 months1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
1.2 7 - 10 months1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
2 Change in FEV1 (L) from baseline3 Mean Difference (IV, Fixed, 95% CI)Subtotals only
2.1 At <24 hours116Mean Difference (IV, Fixed, 95% CI)-1.40 [-3.07, 0.27]
2.2 At 1 - 30 days2139Mean Difference (IV, Fixed, 95% CI)0.06 [-0.00, 0.13]
2.3 At 1 - 2 months2138Mean Difference (IV, Fixed, 95% CI)-0.05 [-0.12, 0.02]
3 Change in FEV1 % predicted from baseline2 Mean Difference (IV, Fixed, 95% CI)Subtotals only
3.1 At 1 - 30 days2139Mean Difference (IV, Fixed, 95% CI)1.48 [-0.43, 3.39]
3.2 At 1 - 2 months2138Mean Difference (IV, Fixed, 95% CI)-2.03 [-4.21, 0.14]
4 Change in FVC3 Mean Difference (IV, Fixed, 95% CI)Subtotals only
4.1 At <24 hours116Mean Difference (IV, Fixed, 95% CI)-1.70 [-3.27, -0.13]
4.2 At 1 - 30 days2139Mean Difference (IV, Fixed, 95% CI)0.02 [-0.05, 0.09]
4.3 At 1 - 2 months2138Mean Difference (IV, Fixed, 95% CI)-0.06 [-0.13, 0.02]
5 Number of inpatient episodes1 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
5.1 Within 5 - 6 months1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
6 Adverse events3 Risk Ratio (M-H, Fixed, 95% CI)Subtotals only
6.1 Mild airway symptoms (Alton-pooled: Cough, wheeze, tight chest)116Risk Ratio (M-H, Fixed, 95% CI)1.0 [0.57, 1.76]
6.2 Rhinitis2139Risk Ratio (M-H, Fixed, 95% CI)0.88 [0.67, 1.14]
6.3 Pharyngitis2139Risk Ratio (M-H, Fixed, 95% CI)1.14 [0.76, 1.70]
6.4 Headache2139Risk Ratio (M-H, Fixed, 95% CI)0.86 [0.51, 1.45]
6.5 Sinusitis2139Risk Ratio (M-H, Fixed, 95% CI)0.90 [0.53, 1.54]
6.6 Moderate: Influenza type symptoms (Alton: pooled)116Risk Ratio (M-H, Fixed, 95% CI)7.0 [1.10, 44.61]
6.7 Abdominal pain2139Risk Ratio (M-H, Fixed, 95% CI)0.66 [0.36, 1.20]
6.8 Asthma2139Risk Ratio (M-H, Fixed, 95% CI)1.94 [0.48, 7.91]
6.9 Chest pain2139Risk Ratio (M-H, Fixed, 95% CI)0.89 [0.50, 1.57]
6.10 Cough137Risk Ratio (M-H, Fixed, 95% CI)6.0 [0.33, 108.56]
6.11 Increased cough2139Risk Ratio (M-H, Fixed, 95% CI)0.88 [0.71, 1.10]
6.12 Dyspnoea2139Risk Ratio (M-H, Fixed, 95% CI)1.03 [0.38, 2.84]
6.13 Fatigue137Risk Ratio (M-H, Fixed, 95% CI)0.97 [0.44, 2.12]
6.14 Fever2139Risk Ratio (M-H, Fixed, 95% CI)1.36 [0.79, 2.36]
6.15 Decreased lung function2139Risk Ratio (M-H, Fixed, 95% CI)1.17 [0.63, 2.18]
6.16 Increased Sputum2139Risk Ratio (M-H, Fixed, 95% CI)0.86 [0.61, 1.20]
6.17 Severe (Alton: pooled)116Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
6.18 CF lung exacerbation2139Risk Ratio (M-H, Fixed, 95% CI)1.40 [0.75, 2.61]
6.19 Hemoptysis2139Risk Ratio (M-H, Fixed, 95% CI)0.88 [0.32, 2.47]
6.20 Lung disorder137Risk Ratio (M-H, Fixed, 95% CI)1.28 [0.43, 3.78]
7 Lower airway potential difference change from baseline1 Mean Difference (IV, Fixed, 95% CI)Totals not selected
7.1 Day 1 - 301 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
8 Lower airway potential difference change from baseline (amiloride and low chloride)1 Mean Difference (IV, Fixed, 95% CI)Subtotals only
8.1 Response to perfusion with amiloride116Mean Difference (IV, Fixed, 95% CI)3.90 [-9.76, 17.56]
8.2 Response to perfusion with zero chloride solution116Mean Difference (IV, Fixed, 95% CI)6.86 [3.77, 9.95]
9 Measurement of CFTR protein expression (SPQ chloride efflux)1 Mean Difference (IV, Fixed, 95% CI)Subtotals only
10 Change in validated computerised tomogram (CT) score1 Mean Difference (IV, Fixed, 95% CI)Subtotals only
Analysis 1.1.

Comparison 1 CFTR gene replacement therapy versus placebo, Outcome 1 Respiratory exacerbations (episodes).

Analysis 1.2.

Comparison 1 CFTR gene replacement therapy versus placebo, Outcome 2 Change in FEV1 (L) from baseline.

Analysis 1.3.

Comparison 1 CFTR gene replacement therapy versus placebo, Outcome 3 Change in FEV1 % predicted from baseline.

Analysis 1.4.

Comparison 1 CFTR gene replacement therapy versus placebo, Outcome 4 Change in FVC.

Analysis 1.5.

Comparison 1 CFTR gene replacement therapy versus placebo, Outcome 5 Number of inpatient episodes.

Analysis 1.6.

Comparison 1 CFTR gene replacement therapy versus placebo, Outcome 6 Adverse events.

Analysis 1.7.

Comparison 1 CFTR gene replacement therapy versus placebo, Outcome 7 Lower airway potential difference change from baseline.

Analysis 1.8.

Comparison 1 CFTR gene replacement therapy versus placebo, Outcome 8 Lower airway potential difference change from baseline (amiloride and low chloride).

Analysis 1.9.

Comparison 1 CFTR gene replacement therapy versus placebo, Outcome 9 Measurement of CFTR protein expression (SPQ chloride efflux).

Analysis 1.10.

Comparison 1 CFTR gene replacement therapy versus placebo, Outcome 10 Change in validated computerised tomogram (CT) score.

Appendices

Appendix 1. Search strategy: National Institutes for Health (NIH) Genetic modification Clinical Research Information System (GeMCRIS)

Search dates: 1992 to 04 September 2013
Search term: "cystic fibrosis" (as the medical condition)

What's new

DateEventDescription
19 September 2013New citation required but conclusions have not changedWhile a new study has been identified, it is still ongoing and has not yet been assessed for inclusion in this review, hence our conclusions remain the same.
19 September 2013New search has been performedA search of the Cystic Fibrosis & Genetic Disorders Group's Cystic Fibrosis Trials Register identified four references to a single ongoing study which is potentially eligible for inclusion in this review, but will be assessed in full once completed (Alton 2012).

History

Protocol first published: Issue 1, 2006
Review first published: Issue 2, 2007

DateEventDescription
29 August 2012New search has been performedA search of the Group's Cystic Fibrosis Trials Register did not identify any new studies potentially eligible for inclusion in the review.
29 August 2012New citation required but conclusions have not changedNo new studies were included in this update of the review, so the conclusions remain the same.
28 September 2011New search has been performed

A search of the Group's Cystic Fibrosis Trials Register and an additional search of GeMCRIS did not identify any potentially eligible references for this review.

A study previously listed as ongoing, has now been excluded after confirmation from the investigators that the study did not have a control arm (Alton 2009).

12 August 2010New search has been performedA search of the Group's Cystic Fibrosis Trials Register identified one additional reference for possible inclusion in this review (Griesenbach 2008). This abstract described plans for a future clinical trial and has not been included.
9 September 2009New search has been performed

A search of the Group's Cystic Fibrosis Trials Register did not identify any potentially eligible references for inclusion in this review.

Through personal communication we have identified a new ongoing trial (Alton 2009a).

12 August 2009AmendedContact details updated.
11 November 2008New search has been performedA search of the Group's Cystic Fibrosis Trials Register did not identify any potentially eligible references for this review.
11 November 2008AmendedConverted to new review format.
20 February 2008New search has been performedA search of the Group's Cystic Fibrosis Trials Register identified one new reference (Moss 2007). This is the full paper relating to the abstracts previously included under the study ID Moss 2005. This paper reported full adverse event follow up that had previously been reported as "incomplete", and the new data have been incorporated into the adverse event analyses in this update.
20 February 2008AmendedWe have replaced the original 'Synopsis' with a new 'Plain language summary' in line with latest guidance from The Cochrane Collaboration.
21 February 2007New citation required and conclusions have changedSubstantive amendment

Contributions of authors

The protocol and review were conceived and drafted by Dr Tim Lee and Dr Kevin Southern. Both authors selected studies and extracted data.

Dr Lee acts as guarantor for the review.

Declarations of interest

None known.

Characteristics of studies

Characteristics of included studies [ordered by study ID]

Alton 1999

Methods

Double-blind randomised placebo-controlled trial, parallel design, intention-to-treat basis.

Single centre in UK.

Participants

16 participants (all male). Treatment group n = 8, placebo group n = 8.

Mean age 26.9 years. Mean (SE) age: treatment group 27·5 (3·4) years, placebo group 26·3 (1·7) years.
Confirmed CF, FEV1 >70%, sterile.

Genotype (not split by treatment/placebo group):

∆F508/∆F508 n = 12

∆F508/W1282X n = 1

∆F508/other (ie, no mutation detected after screening for mutations present in 92–94% of UK patients with CF) n = 3.

InterventionsSingle dose of CFTR DNA+liposome, or liposome alone nebulised to lungs.
OutcomesAdverse events, gene expression, CFTR protein expression, airway potential difference.
NotesAdditional data requested from author, no additional data available.
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskDescribed as randomised, but no further details given.
Allocation concealment (selection bias)Unclear riskNot discussed.
Blinding (performance bias and detection bias)
All outcomes
Low riskParticipants and outcome assessors blinded.
Incomplete outcome data (attrition bias)
All outcomes
High riskAll randomised participants assessed on intention-to-treat basis, although data incomplete for 4 reported endpoints.
Selective reporting (reporting bias)Unclear riskUnclear.
Other biasUnclear riskUnclear.

Moss 2004

Methods

Double-blind randomised placebo controlled trial, parallel design, intention-to-treat basis.

Multicentre - 8 CF centres in the USA.

Participants

37 participants (15 male, 22 female).

Treatment group n = 20 (6 male, 14 female), placebo group n = 17 (9 male, 8 female).
Mean age 23.7 years. Mean (SD) age: in treatment group 24.2 (8.7) years; in placebo group 23.2(11.1) years.
Confirmed CF, FEV1 >60%. Mean (SD) FEV1: in treatment group 82.2 (19.3) and in placebo group 84.4 (15.1),P = 0.70.

Genotype:

F508 homozygous: treatment group 5 (25%), placebo group 13 (77%).

F508 heterozygous: treatment group 12 (60%), placebo group 2 (12%).

Other/unknown: treatment group 3 (15%), placebo group 2 (12%).

Interventions1x10(13) particles tgAAVCF 3 times (30 day interval) or matching placebo, nebulised to lungs.
OutcomesRespiratory exacerbations, adverse events, lung function, inpatient episodes, acquisition of new pathogens, gene expression, change in CT score.
Notes

Additional data requested from author, original data provided.

Power calculation undertaken - paper states " Enrollment of 18 subjects per treatment group provided adequate power to test differences between treatment groups with respect to the primary and secondary protocol end points "

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskDescribed as randomised, but no further details given.
Allocation concealment (selection bias)Unclear riskNot discussed.
Blinding (performance bias and detection bias)
All outcomes
Low riskDouble-blind, no further details.
Incomplete outcome data (attrition bias)
All outcomes
High risk

42 participants randomised, 37 received at least one dose of study drug, 35 completed all 3 doses.
Not clear if withdrawals in placebo or active group, data pooled for 37 participants receiving at least 1 dose. Only 35 participants underwent HRCT lung scans, and just 10 had vector-specific CFTR gene expression assessed.

The first participant, who had received one dose of study medication, was withdrawn from the study as a result of a FDA hold on the research. The second participant was withdrawn prior to dosing, then later was re-screened and was re-randomized, but withdrew consent prior to dosing. 3 other participants withdrew consent after randomization but prior to starting treatment because of heavy work schedule, concern about potential risks, and a change of mind, respectively.

Selective reporting (reporting bias)Unclear riskUnclear.
Other biasHigh riskThe placebo group had significantly more ΔF508 homozygous participants (77%) than the CFTR gene replacement group (25%) (P = 0.01).

Moss 2007

  1. a

    CF: cystic fibrosis
    CFTR: cystic fibrosis transmembrane conductance regulator
    CT: computerised tomogram scan (of chest)
    DNA: deoxyribonucleic acid
    FEV1: forced expiratory volume in one second
    tgAAVCF: Targeted Genetics Corporation adeno-associated virus encoding CFTR

Methods

Double-blind randomised placebo controlled trial, parallel design, intention-to-treat basis.

Multicentre - 12 centres in USA.

Participants

122 people screened, 109 randomised and 102 participants (54 male, 48 female) received treatment. 98 completed study and 4 stopped early (reasons given - see below).

Treatment group n = 51 (26 male), placebo group n = 51 (28 male).
Mean age 22.6 years, all aged over 12 years. Mean (SD) age: treatment group 23.9 (10.9) years, placebo group 21.3 (8.7) years.
Confirmed CF, FEV1 >60% predicted. Mean (SD) FEV1 % predicted: treatment group 84.7 (13.7), placebo group 87.9 (15.5).

Genotype:

AF508 homozygous: treatment group 27 (53%), placebo group 27 (53%).

AF508 heterozygous: treatment group 18 (35%), placebo group 20 (39%).

Interventions1x10(13) particles tgAAVCF 2 times (30 day interval) or matching placebo, nebulised to lungs.
OutcomesRespiratory exacerbations, adverse events, lung function.
Notes

Additional data requested from author, original data provided.

Sample size calculation based on the initial phase 2 multidose aerosol study. "Enrollment of 100 subjects, 50 in each treatment arm, would provide 93% power to detect a 0.14- liter difference in the 30-day change in FEV1 between treatment groups, assuming a standard deviation of the change of 0.20 liter in each group. This sample size would also provide adequate power for detecting 0.3 log10-ng/ml differences between treatment groups in IL-8."

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskDescribed as randomised, but no further details given.
Allocation concealment (selection bias)Unclear riskNot discussed.
Blinding (performance bias and detection bias)
All outcomes
Low riskDouble-blind, no further details.
Incomplete outcome data (attrition bias)
All outcomes
Low risk109 participants randomised, of whom 102 received at least 1 dose of study drug, 98 completed trial - 4 participants stopped early. Of these, 1 was in the tgAAVCF treatment group and 3 were in the placebo group. The reason the participant from the tgAAVCF group discontinued was loss to follow-up. Reasons for discontinuation in the 3 placebo recipients included experiencing an adverse event (unlikely to be related pulmonary exacerbation), death (unrelated motorcycle accident), and other (no response to day 210 phone call).
Selective reporting (reporting bias)High riskSome relevant secondary endpoints, such as number of days of oral antibiotic use, remain unreported
Other biasUnclear riskUnclear.

Characteristics of excluded studies [ordered by study ID]

StudyReason for exclusion
Davies 2011Not a randomised controlled trial
Flotte 1996Not a randomised controlled trial
Gill 1997Applied to nose not to lungs
Harvey 1999Not a randomised controlled trial
Hyde 2000Applied to nose not to lungs
Joseph 2001Not a randomised controlled trial
Knowles 1995Applied to nose not to lungs
Noone 2000Not a randomised controlled trial
Porteous 1997Applied to nose not to lungs
Wagner 1999Applied to sinuses not lungs
Wagner 2002Applied to sinuses not lungs
Zabner 1996Not randomised controlled trial
Zabner 1997Not randomised controlled trial
Zuckerman 1999Not a randomised controlled trial

Characteristics of ongoing studies [ordered by study ID]

Alton 2012

  1. a

    CF: cystic fibrosis
    HRCT: high resolution computer tomography

Trial name or titleA randomised, double-blind, placebo-controlled phase 2B clinical trial of repeated application of gene therapy in patients with CF.
MethodsDouble-blind, randomised, placebo-controlled, parallel design.
ParticipantsApproximately 130 individuals with confirmed CF aged 12 years and over.
InterventionsAerosolised delivery of pGM169/GL67A, or placebo (0.9% saline) to lungs, 12 doses per patient, 4 weeks apart over 48 weeks.
OutcomesLung function, lung clearance index, HRCT score, respiratory exacerbations, safety.
Starting date2013.
Contact informationProf Eric Alton (e.alton@imperial.ac.uk), Professor of Gene Therapy, Imperial College, London, UK.
NotesLikely to report late 2014 or early 2015.

Ancillary