Nintedanib for treating progressive fibrosing interstitial lung diseases excluding idiopathic pulmonary fibrosis

Progressive fibrosing interstitial lung disease is a debilitating condition that increases the risk of dying

3.1 Interstitial lung disease (ILD) encompasses a large and varied group of lung disorders characterised by inflammation and fibrosis of the lung parenchyma. Some ILDs become progressive fibrosing ILD (PF-ILD). Idiopathic pulmonary fibrosis is one of the most common types of ILD and is associated with early mortality (Cottin et al. 2019). One patient expert explained that he understood that most people with cancer have a better life expectancy than people with PF-ILD; although the committee was not presented with these data, the committee appreciated the increased risk of dying and the need for treatment options. Idiopathic pulmonary fibrosis may share some features of natural history with other types of PF-ILD. This appraisal addresses PF-ILD excluding idiopathic pulmonary fibrosis, for which NICE has recommended nintedanib in NICE's technology appraisal guidance on nintedanib for treating idiopathic pulmonary fibrosis (from now on referred to as TA379). Other types of PF-ILD include idiopathic interstitial pneumonias, autoimmune ILDs, hypersensitivity pneumonitis, and sarcoidosis. People with PF-ILD may also have underlying conditions such as rheumatoid arthritis. Clinical experts explained that some underlying diseases start with more inflammation than others, while some mainly have fibrosis (for example, asbestosis) but all end up with fibrotic lungs. The progressive phenotype in PF-ILD is characterised by a gradual decline in lung function, dyspnoea (breathlessness), worsening of physical performance and quality of life, as well as poor response to immunomodulatory therapies and early mortality. The patient expert explained that in some people symptoms evolve rapidly and, within a short time, breathlessness limits physical activity. Ultimately, some people become housebound and dependent on supplementary oxygen before dying within 3 to 4 years after being diagnosed. The committee concluded that PF-ILD is a debilitating condition that increases the risk of dying relative to the general population.

NHS clinicians use multiple criteria to identify and diagnose PF-ILD

3.2 Clinical experts explained that, in the NHS, people with PF-ILD can be seen by different medical specialists, particularly respiratory and rheumatology teams. Clinicians identify disease progression in people with ILD using multiple criteria. These include a relative decline of at least 10% on spirometry in forced vital capacity percentage (FVC%) predicted, based on algorithms that are adjusted for age, sex and height. A worsening of fibrosis (shown on CT scan) and an increase in dyspnoea are also used. Clinical experts explained that these criteria are comparable to the diagnostic criteria of PF-ILD in INBUILD, the company's main trial (see section 3.5). This trial included patients who met at least 1 of the following criteria in the past 24 months, despite treatment with medications used in clinical practice to treat ILD:

A decline in FVC of at least 10% predicted defines progression and is associated with disease deterioration and mortality in PF-ILD

3.3 The committee was aware that a relative (Cottin et al. 2019) or absolute (Goos et al. 2021; Wong et al. 2020) decline in FVC of at least 10% predicted is one of the criteria that defines disease progression in PF-ILD. The company used an absolute decline in FVC% predicted of 10% in its economic model to categorise disease progression and define health states (see section 3.10). The committee also understood that a decline in FVC, measured either as the change from baseline in millilitres or as a percentage of the predicted value, is commonly used in clinical studies as an end point to measure disease progression in ILD. Evidence, confirmed by clinical experts, suggests that a decline in FVC% predicted is associated with disease deterioration and mortality in idiopathic pulmonary fibrosis (Cottin et al. 2019). Clinical experts explained that usually a lower FVC level is associated with a higher risk of death in PF-ILD. The committee was aware that the end point of the key trial was absolute FVC in millilitres, whereas the company modelled FVC% predicted. The committee was not presented with the algorithm chosen by the company to estimate FVC% predicted. The committee concluded that a relative or absolute decline of at least 10% predicted in FVC is an acceptable criterion for disease progression reflecting deterioration in lung function and is associated with higher risk of death in PF-ILD. The committee further concluded it would like to see how the company transformed trial data into modelled data.

Company's positioning of nintedanib in the treatment pathway is appropriate and would reflect an add-on to standard care

3.4 Current treatment of PF-ILD in the NHS includes treatment for the underlying disease (see section 3.1). The conventional treatments used include, but are not limited to, systemic corticosteroids, mycophenolate mofetil, azathioprine, cyclophosphamide, methotrexate, rituximab, infliximab and best supportive care. The clinical experts highlighted that immunosuppressants are commonly used for underlying autoimmune diseases. The company explained that these treatments that are used as standard care do not have a marketing authorisation for PF-ILD. However, the committee was aware that NICE's guide to the methods of technology appraisal (2013), permits it to consider unlicensed treatments in appraisals if their use in clinical practice is established. The clinical and patient experts noted that treatments currently used for PF-ILD treat the inflammation rather than the fibrosis. They explained that there are no randomised placebo-controlled trials that suggest that these treatments delay progression of chronic fibrosis in PF-ILD. The committee heard from clinical experts that sometimes conventional treatments are stopped if they do not slow the decline of pulmonary function, but may not be stopped because they are treating the underlying condition. The committee appreciated that if recommended, nintedanib would be an add-on therapy to conventional treatments. The committee understood it would be offered to people with ILD in whom the underlying disease has progressed despite receiving conventional treatments. The committee concluded that the company's positioning of nintedanib, as an add-on treatment to standard care in people with ILD whose disease has progressed despite receiving conventional treatments, is appropriate.

The baseline characteristics of the INBUILD trial population, with the exception of concurrent treatments, are generalisable to the NHS

3.5 INBUILD was a phase 3, randomised, multicentre trial comparing nintedanib (n=332) with placebo (n=331) in people with PF-ILD (see section 3.2 for diagnostic criteria). The trial consisted of 2 parts: PART A, in which all patients took nintedanib or placebo for an initial period of at least 52 weeks; and Part B, in which patients continued to have blinded, randomised treatment beyond week 52. The trial continued until all patients had completed Part A (reached their week-52 visit) and a benefit and risk profile of nintedanib over the 52 weeks had been assessed, or until a withdrawal. Patients had different treatment periods in Part B. The trial asked those who stopped the treatment to attend all visits as originally planned, including an end-of-treatment visit and a follow-up visit 4 weeks later. There were 2 database locks in the trial. Database lock 1 was beyond week 52, and database lock 2 (median follow up about 19 months, Brown et al. 2020) took place after all patients had completed the follow-up visit. Patients could receive open-label nintedanib in an extension study known as INBUILD-ON (n=436, completed) if the benefit and risk assessment of nintedanib was positive. The committee was aware that INBUILD included 22 patients from the UK. In both arms, patients had largely balanced baseline characteristics for sex, age, and decline in lung function measured before the trial. Clinical experts noted that the criteria for disease progression for entry into the trial, for example, a relative decline in FVC of at least 10% predicted in the 24 months before screening, are in line with what would be seen in practice (see section 3.2). The committee agreed that the baseline characteristics of the INBUILD trial population, with the exception of concurrent treatments (section 3.6), were representative of NHS patients.

Concurrent treatments in the INBUILD trial do not reflect current NHS care

3.6 The committee discussed that in both arms of the INBUILD trial, patients could not have immunosuppressants other than systemic corticosteroids at a maximum dose of 20 mg per day (see section 3.4) at randomisation and for the first 6 months (26 weeks) of the trial. The committee was aware that 68.6% of patients (70.1% in the placebo arm and 67.2% in the nintedanib arm) used corticosteroids over the 52-week period in INBUIILD. Patients who had received other immunosuppressants could participate in the trial if they underwent a wash-out period 4 to 8 weeks before randomisation. After the first 6 months and for the rest of the trial, patients with worsening ILD or connective tissue disease could receive immunosuppressants. Approximately 16% of patients started immunosuppressants during the second 6 months of the initial 52-week period (21% in the placebo arm and 11% in the nintedanib arm). The committee interpreted this to show that fewer patients randomised to nintedanib than placebo needed immunosuppressants, but that a substantial proportion of participants needed the treatments that the protocol restricted earlier in the trial. The ERG noted that immunosuppressants are not restricted in clinical practice and that, because nintedanib is considered to be an add-on treatment (see section 3.4), placebo without conventional standard treatments does not reflect NHS clinical practice. Therefore, it is not an appropriate comparator. The clinical experts explained that nintedanib would be offered to reduce the dosage and use of corticosteroids, but the committee was not presented with any evidence for this. The committee concluded that the INBUILD trial does not represent NHS clinical practice because it restricted immunosuppressant use, and if nintedanib was recommended, it would be added on to current treatments.

Nintedanib is associated with a slower decline of lung function compared with placebo, but its long-term treatment effect is uncertain

3.7 The primary end point of INBUILD (see section 3.5) was annual rate of decline in FVC (in millilitres per year) over 52 weeks adjusted for baseline FVC (in millilitres) and imaging pattern. The committee noted that this is different from FVC% predicted, which defines disease progression in the INBUILD inclusion criteria and as an outcome in the company's economic model. In the trial the company analysed the change in absolute FVC measured in millilitres using a random coefficient regression model with baseline FVC and imaging pattern as covariates. The slope of the FVC decline was calculated for every patient over 52 weeks and the average compared between the 2 arms. The analysis was done over 52 weeks and included people who had discontinued (see section 3.5). Results from adjusted annual rate of decline in FVC showed that:

Nintedanib's treatment effect on mortality is uncertain

3.8 Secondary endpoints assessed in INBUILD included overall survival, change from baseline King's Brief Interstitial Lung Disease (K-BILD) questionnaire total score (a health-related quality-of-life measure) and time until first acute ILD exacerbation or death. Other endpoints included the EQ-5D questionnaire and measures of safety. Endpoints were collected at 52 weeks and at database lock 2 (median follow up: about 19 months, see section 3.5). Results showed that:

The adverse event profile of nintedanib is acceptable

3.9 In INBUILD at 52 weeks, the nintedanib and placebo arms showed similar frequencies of any adverse event and of serious adverse events. Gastrointestinal discomfort, especially diarrhoea, was the most common adverse event (66.9% in the nintedanib arm compared with 23.9% in the placebo arm). People taking nintedanib were more likely to show biochemical indication of hepatic injury and were more likely to reduce the dose of nintedanib because of gastrointestinal symptoms. Based on their observations, the clinical experts explained that 25% to 30% of people may not tolerate nintedanib longer term. They noted that taking nintedanib is not associated with an increased risk of infection unlike some immunosuppressants. The committee concluded that the safety profile of nintedanib was acceptable.

The model structure is acceptable for decision making but there are important uncertainties in its assumptions

3.10 To model the natural history of PF-ILD beyond the end of the trial, the company assumed that idiopathic pulmonary fibrosis and PF-ILD have the same natural history of disease. The company therefore adopted the same Markov model structure it had used in TA379. In the model, patients accrue quality-adjusted life years by improving both quality and length of life. Nintedanib improves the quality of life because patients experience fewer exacerbations, and their lung function declines more slowly. In the model, patients on nintedanib also live longer compared with those on placebo. The company defined health states using lung function (10-point categorisations of FVC% predicted, distribution calculated from the INBUILD trial) and rates of exacerbation over a cycle length of 3 months in the model. Patients enter the model at different lung-function health states without exacerbation. They can then either remain in the same health state, or transition to a health state with the same lung function with the occurrence of exacerbation or with a 10-point lower FVC% predicted with or without exacerbation, or death. This was to reflect a clinically meaningful decline in FVC% predicted (see section 3.7) and the distribution of patients at randomisation in INBUILD.

3.11 The transition probabilities for exacerbation in the model were informed by data from the INBUILD trial's second database lock. The company assumed that patients could not transition to a health state with better lung function. Also, after an exacerbation, patients could not transition to a health state without exacerbation for the rest of the time horizon. Transition to death can either happen from any health state based on a survival analysis using the INBUILD data or, by reaching a level of FVC% predicted below 40%, at which the company assumed that lung function is unsustainable. The committee was aware that the model did not include absolute decline in FVC controlled for baseline FVC and imaging, which defined the primary end point of the trial. The ERG noted that the company modelled mortality based only on overall survival data from INBUILD. This means that mortality is independently modelled from lung function and that the same risk of death is applied to all health states, even for patients with the lowest level of lung function. The committee recalled that a lower level of lung function was strongly associated with a higher mortality rate (see section 3.3) and questioned the company on why it did not model a change in risk of death for lower levels of lung function. The company explained that this was to avoid double counting as the overall survival data includes all deaths. The company also modelled mortality independently from acute exacerbations, despite it stating that acute exacerbations are often fatal and a major cause of mortality in PF-ILD.

3.12 The ERG did not make any change to the company's modelling assumptions but noted that modelling mortality independently from lung-function decline and acute exacerbations can produce implausible results in relation to stopping treatment in the model. The committee understood that the way the company modelled mortality meant that modifying assumptions around the rate at which people stopped treatment with nintedanib or placebo (see section 3.23), exacerbation rates (see section 3.19), and decline in lung function (see section 3.20) would have minimal impact on the cost effectiveness. The committee was aware that mortality data from INBUILD trial were sparse and showed no difference in mortality between treatments (see section 3.8). Considering the trade-off between scarcity in overall survival data and the potential confounding factors in the relationship between level of FVC% predicted, exacerbations, and mortality, the committee concluded that there are important uncertainties in the model structure and limitations in its implementation, which the committee accounted for in its decision making.

The company used the frequentist and Bayesian approaches to extrapolate overall survival

The company used 2 approaches to extrapolate overall survival beyond the trial duration: a frequentist method and an exploratory Bayesian analysis.

In the absence of data, it is reasonable to consider that idiopathic pulmonary fibrosis and PF-ILD share similar disease trajectories including death

3.15 The company's Bayesian approach assumed that PF-ILD and idiopathic pulmonary fibrosis have similar disease trajectories including death (see section 3.14). The committee was aware that PF-ILD and idiopathic pulmonary fibrosis may share similar features and natural history (see section 3.1) and questioned how valid it was to assume similar survival between the 2 populations. The company explained that they based this assumption on 2 sources of data. First, Brown et al. (2020), which is an observational study assessing the clinical course of PF-ILD and idiopathic pulmonary fibrosis, and includes people receiving placebo in the INBUILD trial (PF-ILD) and INPULSIS trials (idiopathic pulmonary fibrosis). The study, sponsored by the manufacturer, reported that PF-ILD and idiopathic pulmonary fibrosis had similar clinical course including death because there were no statistically significant difference in death between the 2 populations at 52 weeks. However, the committee was aware that the HR of 0.63 (95% CI 0.35 to 1.13) suggested the possibility of a lower rate of death for PF-ILD than for idiopathic pulmonary fibrosis. Given that deaths were sparse in people taking placebo in both the INBUILD trial (5.1%; 17 out of 331) and INPULSIS trials (7.8%; 33 out of 423), and the lack of reporting of what known or unknown confounding factors the authors adjusted for in Brown et al., the committee considered that the evidence should be interpreted with caution. Simpson et al. (2020), an observational study of PF-ILD across England, included 2,368 people with a new referral for ILD diagnosed with the same criteria for PF-ILD as used in the INBUILD trial (see section 3.2). The study reports that people with PF-ILD and people with idiopathic pulmonary fibrosis have similar survival at 900 days (HR 1.06; 95% CI 0.84 to 1.35). Additionally, it showed that the shapes of the survival curves between PF-ILD and idiopathic pulmonary fibrosis were similar. The committee considered that the shape of the survival curve was more important than the actual HR when assuming similar disease trajectories, particularly given the approach taken by the company in its Bayesian analysis. The committee also recalled that idiopathic pulmonary fibrosis is a subset of PF-ILD (see section 3.1) and agreed it was reasonable to assume similar survival between people with PF-ILD and idiopathic pulmonary fibrosis. It concluded that the assumption that people with idiopathic pulmonary fibrosis and PF-ILD have similar disease trajectories, when controlling for confounders, on which the company based its Bayesian approach, is reasonable.

Modelling and validation of overall survival for the placebo arm is uncertain and its impact on the model results is not clear

3.16 The committee discussed the company's validation of the Weibull Bayesian survival curves for both nintedanib and placebo arms. It noted that the company chose the Bayesian Weibull curves for both nintedanib and placebo based on statistical fits, clinician input and comparison with external idiopathic pulmonary fibrosis registry data. Clinicians consulted by the company noted that there is limited knowledge on nintedanib's long-term impact. For the placebo arm, the clinicians consulted by the company agreed that Weibull curves (both frequentist and Bayesian) were clinically plausible. The company also used several idiopathic pulmonary fibrosis registries to validate the survival curves fitted for the PF-ILD population in the Bayesian approach, specifically:

3.17 Considering the Bayesian approach itself reasonable (see section 3.14), the committee noted the following with regards to the company's validation of survival curves:

There are uncertainties in fitting independent parametric survival distributions to both arms and the observed and implied hazards should be explored

3.18 The company fitted parametric distributions to the nintedanib and placebo arms independently to extrapolate overall survival beyond the trial data in both its frequentist and Bayesian approaches. The committee understood that when parametric distributions are fitted individually to each treatment arm, no assumptions about the proportionality of treatment effects are made. The committee noted that the company's chosen fits resulted in ever-increasing survival benefits for nintedanib compared with the placebo in the extrapolated portion of the survival curves. The committee considered that this is not supported by the short-term evidence on mortality from the INBUILD trial (see section 3.8) and recalled that the evidence suggested that nintedanib's treatment effect may decrease in the longer term (see section 3.7). The committee therefore considered that the company's approach of fitting parametric survival curves individually to the nintedanib and placebo arms may bias estimates of cost effectiveness in favour of nintedanib. The committee was not provided with evidence that the company had explored the proportionality of treatment effects in the observed data and had not been presented with information on the treatment effect over time implied by the company's chosen curves. It concluded that the company should explore the proportionality of hazards observed in the data and provide information on the treatment effect implied by the alternative survival modelling approaches considered.

The distribution used to model acute exacerbations is acceptable

3.19 The company modelled time to first acute exacerbation using an exponential curve that generated a per-cycle (3-month cycle length) risk of exacerbation of 1.76% and 1.12% for people receiving placebo and nintedanib, respectively. The ERG noted that the exponential model overpredicted the risk of acute exacerbation after approximately 8 months and showed a sharp drop towards the end of follow up. This is likely to substantially increase the difference observed between nintedanib and placebo. The committee was aware that both the company and the ERG's varying risk of exacerbation in scenario analyses had little impact on the cost effectiveness. This was likely because of mortality not being directly linked to the occurrence of acute exacerbation in the model (see section 3.12), representing an important limitation of the model. The committee concluded that the curve used to model acute exacerbations was acceptable, but that the lack of a link between exacerbations and mortality was an important limitation of the model.

The modelling of loss in lung function is acceptable but the lack of link to mortality is an important limitation

3.20 Patients started the model in different FVC% predicted categories (distribution informed by patients' baseline FVC% predicted values in the INBUILD trial; see section 3.10). The company used 2 different methods to calculate probabilities of losing lung function in the placebo arm (multivariate logistic regression based on data from INBUILD trial) and nintedanib arm (odds ratio sourced from INBUILD trial applied to baseline placebo risk). At clarification, the ERG asked the company to use the regression analysis for both arms, but this change had a minimal impact on the cost effectiveness. The ERG noted that this may be because that, while the absolute values differ substantially, the relative differences between pre- and post-exacerbation and between nintedanib and placebo do not differ substantially between the 2 methods. It also noted that that both methods assumed a lifetime treatment effect for nintedanib. The committee was aware that decline of lung function is not linked to mortality (see section 3.12) in the model, and that data from the INBUILD trial suggested that the treatment effect of nintedanib may decrease over time. The committee concluded that the method used to model the loss of lung function over time was acceptable but there were uncertainties over what algorithm the company used to estimate FVC% predicted and the treatment effect over the long term. It recognised that the lack of link between loss of lung function and mortality was an important limitation of the model.

The ERG's adjustment of utility values is appropriate

3.21 EQ-5D was collected in INBUILD and the company used this data to estimate the utility values in the model for each FVC% predicted health state (see section 3.10). The regression analysis resulted in a higher value in people in the 80% to 89% FVC% predicted category than the ones in the 90% to 99% predicted category (0.7333 compared with 0.7287). Because the ERG considered this clinically implausible, they corrected it by applying a linear decline in utility from the 90% to 99% predicted and 70% to 79% predicted health states. This resulted in a utility value of 0.7265 for the 80% to 89% FVC% predicted health state in its base case cost-effectiveness analyses. The committee concluded it was satisfied with the ERG's utility adjustment.

The company's choice of utility decrement for recurrent exacerbations and gastrointestinal events is appropriate

3.22 The company modelled utility decrements for recurrent exacerbations and adverse events. When people have an acute exacerbation, their utility drops by 0.167, which the company estimated from regression analysis using the EQ-5D data collected in INBUILD. The ERG identified 2 other estimates for acute exacerbations and explored the impact of those in scenario analyses. When people have a gastrointestinal adverse events, their utility drops by 0.034. The ERG based this on estimates from TA379, and the assumption that nintedanib has a similar safety profile regardless of being used to treat PF-ILD or idiopathic pulmonary fibrosis. In TA379 the company had used a decrement of 0.068. But the company assumed that 0.034 (half of 0.068) was plausible for PF-ILD based on a phase 3 trial in recurrent non-small-cell lung cancer for nintedanib, which estimated a decrement for grade 3/4 diarrhoea of 0.042 (Boehringer Ingelheim: data on file, 2014). These were estimated from EQ-5D data from the INPULSIS trial in idiopathic pulmonary fibrosis. The ERG was unclear on the company's rationale behind the use of 0.034 rather than 0.068. However, it noted that adverse events had limited impact on the model results. The ERG identified 2 other estimates for gastrointestinal events and explored the impact of those in scenario analyses. The committee was broadly satisfied with the company's choice of utility decrements for recurrent exacerbations and gastrointestinal events but noted these had minimal impact on the cost effectiveness.

The way the company modelled time-to-stopping treatment is uncertain and may underestimate the cost of nintedanib

3.23 The company modelled time-to-stopping treatment with nintedanib using an exponential curve. The company did this for consistency with TA379 and it resulted in an overall rate of 5.97% for stopping nintedanib per month. The company considered that the exponential model did not fit the INBUILD trial data. This is because the model underestimated stopping nintedanib in the first year, but from approximately 15 months onwards, the model appeared to overestimate stopping nintedanib. After validating the model with long-term data on safety and efficacy of nintedanib in people with idiopathic pulmonary fibrosis (Lancaster et al. 2019), the company considered that the exponential model may underestimate the true rate of stopping nintedanib, therefore overestimated its costs. However, the committee noted that the exponential curve dropped quickly and seemed to overestimate stopping of nintedanib compared with the INBUILD trial, and therefore underestimated its costs. In response to the technical engagement, the company also provided alternative models to assess the impact of using different curves to extrapolate time-to-stopping treatment with nintedanib beyond the trial period. The Gompertz curve was closest to the Kaplan–Meier data from INBUILD over 3 years, but it unrealistically underestimated the stopping of nintedanib treatment over the long term. The ERG noted that, given the mean age of 65 years at baseline in the INBUILD trial, the Weibull model probably gave a more realistic extrapolation of time on nintedanib over the long term than other curves. This is because most patients would not stay on treatment for 35 years. The committee noted that the choice of distribution does not impact on the quality-adjusted life years and costs, other than the acquisition cost for nintedanib. This is because different health states in the model are not associated with different mortality rates (see section 3.11). The committee concluded that the way the company has modelled time-to-stopping treatment with nintedanib was uncertain and may have underestimated the cost of nintedanib in the model.

The company's submission does not make a case for end of life care

3.24 The committee considered the advice about life-extending treatments for people with a short life expectancy in NICE's guide to the methods of technology appraisal. The company did not make a case for end of life care in its submission. The committee noted that the life expectancy of people who had placebo would normally be longer than 24 months.

The end-of-life criteria are not met

3.25 The evidence suggested that the median overall survival was expected to be 2 to 5 years in people with PF-ILD not receiving anti-fibrotic therapy (Meltzer et al. 2008; Raghu et al. 2011). The committee concluded that the end-of-life criteria were not met for nintedanib in PF-ILD.

The cost effectiveness of nintedanib for PF-ILD is unknown

3.26 Because of confidential commercial arrangements for nintedanib, the cost-effectiveness estimates cannot be reported here. The committee noted that none of the company's or ERG's analyses reflected assumptions that addressed the committee's concerns and, therefore, the true incremental cost-effectiveness ratio was unknown.

Uncertainties should be explored

3.27 The committee noted that the company should explore the following:

Nintedanib does not meet NICE's criteria for an innovative treatment

3.28 The company considered nintedanib to be innovative because it is the only treatment licensed for people with PF-ILD. In addition, the company stated that nintedanib is the first pharmacological treatment to show clinical evidence of slowing disease progression in people with PF-ILD. However, given the shortcomings of the company's modelling, the committee could not determine whether this reflects a 'step change' in treatment.

There are no equality issues within the committee's remit

3.29 The patient experts considered there was inequality because people with idiopathic pulmonary fibrosis can access nintedanib while people with PF-ILD currently cannot. The committee noted that equality issues could only be assessed within the marketing authorisation defined in this appraisal. The patient experts further explained that people with PF-ILD are generally younger than people with idiopathic pulmonary fibrosis and more likely to be of South Asian and African-Caribbean family origin, which the committee noted was not an equality issue itself in this topic. The committee concluded that there were no equality issues within the committee's remit.

The committee cannot assess nintedanib's long-term effectiveness and value for money and therefore it is not recommended

3.30 The committee acknowledged that PF-ILD is a debilitating condition associated with morbidity and mortality. It noted that if recommended, nintedanib would be an add-on to the current standard care. It noted that the clinical evidence suggested that nintedanib may slow the decline of lung function, but there were uncertainties about its effectiveness and around the assumptions and approaches in the model. The committee concluded that: