Thorax Online First, published on March 5, 2013 as 10.1136/thoraxjnl-2011-201268 Editorial Thorax: first published as 10.1136/thoraxjnl-2011-201268 on 5 March 2013. Downloaded from gained compared with triple therapy.4 Pirfenidone should be prescribed While at first glance pirfenidone seems expensive, it is only about twice the price for patients with idiopathic of triple therapy, so maybe it is not quite fi as expensive as previously thought. The pulmonary brosis current threshold for NICE approval is between £20 000 and £30 000 per QALY. Gisli Jenkins The QALY is an estimate of effectiveness based on both quality-of-life, using pri- marily the EuroQal ED-50 questionnaire, Pirfenidone works. There have been four 2008 BTS IPF guidelines9 implicitly as well as life expectancy. It is clear the randomised placebo-control trials of pirfe- recommended treatment with triple QALY is not going to favour pirfenidone, nidone for the treatment of idiopathic pul- therapy, although this advice has been because the clinical trials were not monary fibrosis (IPF); a phase II and phase modified since the publication of the powered to detect changes in mortality, III study in Japan, and two international Panther study.10 The primary justifications nor did they measure the ED-50 scores. – multicentre phase III studies.1 3 In all four for using these therapies were: the possibil- Furthermore, a major limitation of the studies, patients treated with pirfenidone ity of benefit in some patients; low cost; QALY is that it only considers the cost- had slower rates of decline in lung volume and, at least for NAC, lack of major effectiveness of a therapy in relation to (vital capacity (VC) in the Japanese studies adverse effects. the patient taking the drug, excluding and forced vital capacity (FVC) in the Pirfenidone, unlike previously recom- effects on third parties, such as relatives international studies) than placebo, and in mended therapies, has demonstrated sig- and carers, and the global economic bene- three studies, the results were statistically nificant improvements in lung function fits of bringing novel therapies to market. significant. National Institute of Clinical compared with placebo in randomised Are effects on lung physiology worth Excellence (NICE) agrees that pirfenidone controlled trials (RCTs). Furthermore, pir- paying for? Change in FVC is an accepted has a ‘modest but measurable effect on fenidone improves progression-free sur- marker of mortality and disease progres- slowing the decline in lung function’.4 vival,11 6 min walking distance,3 and sion in IPF.14 It is a clinically useful meas- Therefore, whether patients receive pirfe- reduces acute exacerbations of IPF.1 None urement; it was the primary endpoint in nidone depends on whether pirfenidone of the studies were powered to assess mor- the CAPACITY study, but it has little works enough to justify its cost. tality but, overall, 0.75% of patients impact on the QALY. Some argue that all IPF is a chronic progressive disease of (3/406) receiving pirfenidone died during phase III studies in IPF should be unknown aetiology, it is fatal with a the studies compared with 3.8% of powered to detect mortality as the median survival of approximately 3 years,5 patients receiving placebo (12/312). In the primary endpoint.15 While powering clin- or less than one electoral cycle. No recent interim analysis of the Panther ical trials for mortality to demonstrate

therapy has been proven to improve study, 1.3% of patients receiving placebo harm would not require large trials, a http://thorax.bmj.com/ survival. Despite the absence of good (1/77) had died compared with 10.3% of study demonstrating improved survival is evidence, various combinations of patients receiving triple therapy (8/78).12 likely to need nearly 2600 patients fol- N-acetylcysteine (NAC), prednisolone and Pirfenidone causes dermatological (photo- lowed for 5 years.16 This would exponen- azathioprine have been used for many sensitivity) and gastrointestinal (dyspepsia, tially add to clinical trial costs, which years to treat patients with IPF. In the early diarrhoea, vomiting and anorexia) would ultimately have to be passed on to 1990s, a small study suggested that pred- symptoms, and a slight increased risk of the consumer and, thus, reduce the cost- nisolone and azathioprine might be benefi- neurological disturbance (dizziness, fatigue, effectiveness of the drug (because per- 6 cial, and the initial British Thoracic insomnia and anxiety), leading to a forming the trial does not improve effi- on September 23, 2021 by guest. Protected copyright. Society (BTS) guidelines for the treatment 15.7% drop-out rate in patients receiving cacy of a drug, it merely demonstrates of cryptogenic fibrosing alveolitis (as IPF pirfenidone in clinical trials, compared whether the drug has efficacy or not). was then known) recommended this with 8.9% of patients receiving placebo. Therefore, if NICE do not consider therapy.7 In 2005, the Idiopathic So are these effects worth the cost? The change in FVC as an endpoint worth Pulmonary Fibrosis International Group absence of effective therapy suggests the paying for, it means clinical trials in IPF, Exploring N-Acetylcysteine I Annual average cost of treatment for a patient and thus drugs for IPF, will become con- (IFIGENIA) study demonstrated that NAC with IPF should be relatively low: siderably more expensive. might be beneficial in IPF.8 However, this single-agent NAC comes in at just over IPF is the Cinderella disease of the study was criticised at the time for includ- £200/year/patient; treatment is predomin- Cinderella speciality (see table 1), killing ing azathioprine and prednisolone in both antly outpatient based; and high mortality nearly 4000 people in the UK each year; the placebo and treatment arms. The inclu- rates mean treatment duration is short. therefore, more people are dying of IPF sion of this faux placebo lead some, pos- NICE estimates outpatient treatment costs than of many cancers, including cervical, sibly correctly, to suggest that NAC was for IPF with best supportive care at £800 ovarian, pancreatic and renal cancer. inhibiting the adverse effects of prednisol- per year,4 although estimates from claims Within the field of respiratory medicine, one and azathioprine, rather than posses- databases in the USA suggest the cost of the ugly sisters of asthma and chronic sing any direct antifibrotic effect. The IPF closer to US$17 000 for an out- obstructive pulmonary disease (COPD) patient.13 NICE has calculated the incre- take a lion’s share of resources. Seretide is mental cost-effectiveness ratio (ICERs) at the biggest single drug expenditure in the Correspondence to Dr Gisli Jenkins, £36 327 per quality adjusted life years National Health Service (NHS), costing Nottingham Respiratory Research Unit, University of 17 Nottingham, Clinical Sciences Building, City Hospital (QALY) gained, compared with best sup- £366.2 million per year. The NICE Campus, Hucknall Road, Nottingham NG5 1PB, UK; portive care (using the remarkably low guidelines for COPD treatment lay out [email protected] UK estimate), but only £16 560 per QALY the standards for use of combination

CopyrightJenkins G. Thorax ArticleMonth author 2013 Vol 0(or No 0their employer) 2013. Produced by BMJ Publishing Group Ltd (& BTS) under licence.1 Editorial Thorax: first published as 10.1136/thoraxjnl-2011-201268 on 5 March 2013. Downloaded from incrementally effective therapies, 5-year Table 1 Estimates of spending on health care, medical research and professional football survival had reached 85% by the year Activity Annual spend (£) 2009.21 The dramatic improvements seen in Total UK healthcare spending25 121.5 billion cancer outcomes come at a cost. In 2010, Cost of cardiac care to NHS26 15 billion the National Cancer Research Institute in UK spending on non-acute secondary care25 7 billion the UK spent £100 million on breast Cost of respiratory care to NHS27 3 billion cancer research, up from £45 million in Premier League Football wages28 1.6 billion 2002,22 whereas UK academic institutions CRUK research budget22 504 million might possibly spend £1 million per Atorvastatin cost17 322 million annum on IPF research. However, Manchester United wage bill28 153 million Intermune, the worldwide licenced manu- Average Premier League player salary28 1.4 million facturer of pirfenidone, has spent an BLF research fund29 1 million average of $77 million per annum on IPF BLF allocation to IPF29 800 000 (total since 1985) R&D in the last 3 years.23 Pharmaceutical Estimated cost of outpatient IPF care4 800 company interest in antifibrotic therapies IPF, idiopathic pulmonary fibrosis; NHS, National Health Service; CRUK, Cancer Research UK; BLF, British Lung has never been greater, highlighted by an Foundation. editorial in Nature Biotechnology.24 This is because there is widespread belief within the pharmaceutical industry that fibrotic steroid/long acting beta agonist (LABA), and response; further reducing the finan- processes are amendable to therapy, and and long-acting muscarinic antagonist cial burden of pirfenidone. that investment costs can be recovered. (LAMAs) inhalers, but large numbers are The introduction of pirfenidone should This will undoubtedly lead to improve- prescribed outside of these criteria. be the first small step in a long-term ments in patient care. These development However, the unit cost is relatively small, strategy for improving the outlook for costs have to be recovered; the issue is how thus, few questions are asked, and the patients with IPF. The strategic goal we value, and reimburse, these costs. result is a multimillion pound bill to the should be to follow the model proposed Failure to reimburse therapies that have NHS picked up by IPF patients. If every by the National Cancer Research Institute, been shown to be effective could dramatic- patient with IPF in the UK were to receive namely improved patient care through ally hinder market confidence, with pirfenidone, it would cost little more than research. This has generated small incre- knock-on effects to the UK economy, and the annual cost of tiotropium (£113 mental benefits in outcome, rather than future patient care; certainly that is a view 17 million). In the current economic any large paradigm shift, which ultimately, that Germany, France, Italy, Canada and climate of flatlining NHS expenditure, pir- over time, revolutionised outcomes for others seem to have taken. fenidone could be incorporated into many cancers. The last decade has seen the In a disease with a limited evidence http://thorax.bmj.com/ respiratory budgets by guideline-driven, blossoming of such a strategy. At the start base, only one drug, pirfenidone, has been generic prescribing of current inhaled of the new millennium, there had been shown to have a beneficial effect on the therapy before we even have to start four studies that had recruited 114 patients clinical progression of IPF. The only real looking toward big spending specialties into clinical trials. However, since 2000, arguments against the use of pirfenidone such as cardiology. there has been an explosion in the number are its cost and the relevance of change in Not every patient, however, will need of patients with IPF entered into clinical FVC as an endpoint for IPF. However, the to receive pirfenidone. In a genetically trials (3849 patients entered into 15 pub- costs of pirfenidone are not as high as ima-

heterogenous group, such as the UK lished trials). This has been driven, in part, gined when placed in the context of treat- on September 23, 2021 by guest. Protected copyright. population or patients involved in an by the rising incidence of IPF, up by 30% ment for respiratory disease generally, and 18 RCT, there is a graded response from a in the last decade, in part, due to net- could easily be accommodated through small minority of dramatic responders, to works, such as IPFNet in the USA and, in generic prescribing and adherence to the many who do not respond. Those part, due to pharmaceutical interest in guidelines. Furthermore, demonstrating who believe in ‘precision medicine’, fibrosis. This profusion of clinical trials has efficacy against a marker of disease pro- including Lord Darzi, realise that it is the generated huge swathes of data that has gression, such as FVC, in a chronic pro- few, potentially identifiable, ‘responders’ improved our understanding of the natural gressive disease characterised by loss of within the population that skew the treat- history of IPF, exemplified by the recogni- lung volume has the advantage of being ment effect. The same is true of, and also tion of the ‘acute exacerbation’ as the most able to determine response in individual much more obvious for, adverse effects. dramatic and devastating complication of patients avoiding the need to treat This applies to all drugs, and ideally, we IPF, with a 30-day mortality of around all-comers. could predict responders before giving 70%. The trials have also demonstrated Therefore, I believe pirfenidone should therapy on the basis of known mechanism that some of our accepted, if unsubstanti- be offered to patients with mild to moder- of action and stratification against a ated, therapies were, at best, futile and, at ate IPF, who have progressive disease, and 12 19 molecular marker. Unfortunately, pirfeni- worst, harmful. Finally, a few trials show evidence of response at 6 months. done is a ‘good old-fashioned’ drug, and have suggested that specific antifibrotic The effect of denying access to the ONLY 320 nobody really knows how it works. therapies may be effective. IPF research treatment that has ANY efficacy signal in However, because of its known effect on has reached the point that breast cancer an IPF patient population, a group that lung function, what could and should research arrived at in the early 1970s, has been systematically underfunded for happen (and what clinicians have been when patients had the enviable 5-year sur- decades despite a prognosis worse than doing for decades), is that therapy would vival rate of 52% (current 5-year survival most cancers, would devastate the IPF be provisional upon disease progression for IPF is around 25%), but through community.

2 Jenkins G. Thorax Month 2013 Vol 0 No 0 Editorial Thorax: first published as 10.1136/thoraxjnl-2011-201268 on 5 March 2013. Downloaded from Competing interests Consultancy fees received from 7 The Diffuse Parenchymal Lung Disease Group of the treatment trials: the best is the enemy of the good. GlaxoSmithKline, Intermune. Research contracts British Thoracic Society. The diagnosis, assessment Thorax 2012;67:938–40. received from GlaxoSmithKline, Novartis. Lecture fees and treatment of diffuse parenchymal lung diseases 17 http://www.gponline.com/News/article/1027323/Top- received from Boehringer Ingelheim. in adults. Thorax 1999:54;(Suppl I):S1. 10-branded-drugs-NHS-spend/ (accessed Jan 2013). 8 Demedts M, Behr J, Buhl R, et al. High-dose 18 Navaratnam V, Fleming KM, West J, et al. The rising Provenance and peer review Commissioned; acetylcysteine in idiopathic pulmonary fibrosis. N Engl incidence of idiopathic pulmonary fibrosis in the UK. internally peer reviewed. J Med 2005;353:2229–42. Thorax 2011;66:462–7. To cite Jenkins G. Thorax Published Online First: 9 Bradley B, Branley HM, Egan JJ, et al. Interstitial lung 19 Noth I, Anstrom KJ, Calvert SB, et al.A [please include Day Month Year] doi:10.1136/ disease guideline: the British Thoracic Society in placebo-controlled randomized trial of warfarin in thoraxjnl-2011-201268 collaboration with the Thoracic Society of Australia idiopathic pulmonary fibrosis. Am J Respir Crit Care – Thorax 2013;0:1–3. and New Zealand and the Irish Thoracic Society. Med 2012;186:88 95. – fi doi:10.1136/thoraxjnl-2011-201268 Thorax 2008;63(Suppl 5):v1 58. 20 Richeldi L, Costabel U, Selman M, et al.Ef cacy of a 10 http://www.brit-thoracic.org.uk/Guidelines/Interstitial- tyrosine kinase inhibitor in idiopathic pulmonary Lung-Disease-DPLD-Guideline.aspx (accessed fibrosis. N Engl J Med 2011;365:1079–87. REFERENCES Jan 2013). 21 http://www.cancerresearchuk.org/cancer-info/ 1 Azuma A, Nukiwa T, Tsuboi E, et al. Double-blind, 11 Spagnolo P, Del Giovane C, Luppi F, et al. cancerstats/types/breast/survival/breast-cancer- placebo-controlled trial of pirfenidone in patients Non-steroid agents for idiopathic pulmonary survival-statistics#Trends (accessed Jan 2013). with idiopathic pulmonary fibrosis. Am J Respir Crit fibrosis. Cochrane Database Syst Rev 2010;9: 22 http://scienceblog.cancerresearchuk.org/2011/06/29/ Care Med 2005;171:1040–7. CD003134. near-doubling-of-uk-cancer-research-funding- 2 Taniguchi H, Ebina M, Kondoh Y, et al. Pirfenidone 12 Raghu G, Anstrom KJ, King TE Jr, et al. Prednisone, in-less-than-10-years/ (accessed Jan 2013). in idiopathic pulmonary fibrosis. Eur Respir J azathioprine, and N-acetylcysteine for pulmonary 23 http://phx.corporate-ir.net/phoenix.zhtml? 2010;35:821–9. fibrosis. N Engl J Med. 2012;366:1968–77. c=100067&p=irol-reportsAnnual (accessed Jan 2013). 3 Noble PW, Albera C, Bradford WZ, et al. Pirfenidone 13 Collard HR, Ward AJ, Lanes S, et al. Burden of 24 Allison M. Stromedix acquisition signals growing in patients with idiopathic pulmonary fibrosis illness in idiopathic pulmonary fibrosis. J Med Econ interest in fibrosis. Nat Biotechnol. 2012;30:375–6. (CAPACITY): two randomised trials. Lancet 2012;15:829–35. 25 http://www.nuffieldtrust.org.uk/sites/files/nuffield/ 2011;377:1760–9. 14 Raghu G, Collard HR, Egan JJ, et al.Anofficial ATS/ publication/121203_a_decade_of_austerity_ 4 http://guidance.nice.org.uk/TA/Wave23/25/ ERS/JRS/ALAT statement: idiopathic pulmonary summary_1.pdf (accessed Jan 2013). Consultation/DraftGuidance (accessed Jan 2013). fibrosis: evidence-based guidelines for diagnosis and 26 http://www.cks.nhs.uk/cvd_risk_assessment_ 5 Gribbin J, Hubbard RB, Le Jeune I, et al. Incidence management. Am J Respir Crit Care Med and_management/background_information (accessed and mortality of idiopathic pulmonary fibrosis and 2011;183:788–824. Jan 2013). sarcoidosis in the UK. Thorax 2006;61:980–5. 15 Raghu G, Collard HR, Anstrom KJ, et al. Idiopathic 27 http://www.brit-thoracic.org.uk/Delivery-of- 6 Raghu G, Depaso WJ, Cain K, et al. Azathioprine pulmonary fibrosis: clinically meaningful primary Respiratory-Care/Burden-of-Lung-Disease-Reports. combined with prednisone in the treatment of endpoints in phase 3 clinical trials. Am J Respir Crit aspx (accessed Jan 2013). idiopathic pulmonary fibrosis: a prospective Care Med 2012;185:1044–8. 28 Deloitte Annual Report on Football Finance 2012. double-blind, randomized, placebo-controlled clinical 16 Wells AU, Behr J, Costabel U, et al. Hot of the 29 Stevens C. BLF-funded-IPF-research. http://www.blf. trial. Am Rev Respir Dis 1991;144:291–6. breath: mortality as a primary end-point in IPF org.uk/ (accessed Jan 2013). http://thorax.bmj.com/ on September 23, 2021 by guest. Protected copyright.

Jenkins G. Thorax Month 2013 Vol 0 No 0 3