Pharmacologic Treatment of Idiopathic Pulmonary Fibrosis
On the basis of the available RCT data, clinical superiority of one agent over the other has not been established. Both agents have been demonstrated in multiple phase 3 trials to slow the decline in FVC in patients with mild to moderate IPF with acceptable safety profiles. It could be argued that pirfenidone holds the advantage of demonstrating a reduction in all-cause and IPF-related mortality on pooled analysis of its three major trials. On the contrary, nintedanib appears to have an effect in preventing acute exacerbations of IPF, based on central adjudication of events from the INPULSIS pooled patient populations. A recent meta-analysis performed by Loveman et al. indirectly compared the rate change in FVC and reported nintedanib had 'superior benefit' on the rate of change in FVC. It is ill-advised comparing one trial to the next to determine better efficacy of these agents. The ASCEND, CAPACITY, and INPULSIS studies were different in a number of ways. Although all the studies were in well documented IPF, nuanced inclusionary and exclusionary criteria differences resulted in slightly different populations being studied (see Table 4 ). While the primary outcome measure for all studies was the change in FVC, this change was analyzed differently for pirfenidone and nintedanib. Specifically, it was the slope generated from all FVC measures through the course of the study that was used to demonstrate the efficacy of nintedanib, whereas pirfenidone's efficacy was established through a rank analysis of covariance (ANCOVA) analysis. Both the ASCEND and INPULSIS studies were 52 weeks in duration, which seemingly would allow a comparison of change in the FVC between all groups over this unifying period. However, once again, this assumption is fraught with danger and misinterpretation. For example, patients who died in the ASCEND study were imputed a zero value for their FVCs which drove down the rate of change for both the placebo and treatment arms in this study.
Since the efficacy cannot and should not be compared, the decision as to which agent to choose is predicated to a large extent by the potential side effects and patient preference. This is based on cognisance of the patients' prior medical history, their lifestyle, and their 'comfort' zone. Indeed, a salient goal is to maintain patients on therapy since drug discontinuation is more likely to be deleterious to their subsequent outcomes than the choice of agent. Perhaps, as the underlying pathogenesis of IPF becomes more apparent, we will be able to identify likely 'responders' to one agent or another. For now, we assume clinical equipoise and base our decision on the tolerability of the individual medications.
There is no universally recognized system whereby patients' severity of disease is graded. Clinicians typically look at the FVC% predicted and the percent predicted diffusing capacity of the lung for carbon monoxide (DLCO%) predicted (to a lesser extent) to provide a gestalt of this. In terms of the ASCEND and INPULSIS studies, patients with 'mild to moderate' disease were enrolled as defined by the inclusion criteria shown in Table 4 . When the US FDA approved both drugs, their respective prescribing information states that they are approved 'for the treatment of IPF' without any caveat as to the severity. On the one end of the philosophical spectrum, a purist approach would be to limit prescription of these drugs to those IPF patients who meet the pulmonary function test criteria of the respective studies. Arguably then, perhaps patients would need to meet all the inclusion and exclusion criteria of the respective studies. Unfortunately, it is a small minority of IPF patients (~27%) who meet all criteria for enrollment in IPF clinical trials. A more pragmatic approach encompasses the assumption that since the drugs have been shown to work in a select IPF population, they are likely to work in other IPF patients. Given the poor prognosis of IPF and lack of alternative therapies, our practice is to consider and discuss the possibility of therapy in all IPF patients.
On the one end of this spectrum are those patients in whom IPF was diagnosed through an incidental finding and who are asymptomatic with normal lung function. It conceptually makes sense to introduce therapy as early as possible in an effort to preserve pulmonary function for as long as possible. While it might not be unreasonable to observe these patients, we believe that it is nonetheless a discussion to be had with each patient so that they may participate in this decision. Indeed, an asymptomatic patient may present months or years later with symptoms and significant loss of lung function. Since both agents slow the rate of loss of lung function and neither restores lung function, in the absence of a prior documented discussion about therapy, the practitioner might be at medicolegal jeopardy for having withheld medication that might have prevented or slowed the onset of symptoms.
What of those patients with more severe disease, specifically a FVC below 50% predicted? This population of patients are frequently excluded from clinical trials, and it is therefore unknown if oral antifibrotics (OAFs) retain the same efficacy in severe disease as in mild to moderate disease. Limited retrospective data suggest that pirfenidone slows the rate of decline in patients with FVC below 60% predicted. Furthermore, a subgroup analysis from the combined CAPACITY and ASCEND datasets did not show a difference in drug efficacy between patients above and below the median FVC% of the study population. It might be inferred from this that the severity of disease at baseline does not make a difference with regards to response, although, of course, this analysis only included patients from within these studies. Patients with an FVC below 50% of that predicted are not a homogenous group in terms of disease severity. There are those who might be New York Heart Association functional class 2 or even 3 on low flow supplemental oxygen, whereas others are tachypneic at rest on high flow oxygen. In the former situation, therapy might be entirely reasonable, whereas in the latter cases, it might be regarded as futile. Whenever treatment with an OAF is considered and prescribed, it should only occur after a frank discussion of the expectations of therapy, and the possibility and spectrum of side effects of the drugs.
The efficacy of OAFs for longer than 52–72 weeks has not been rigorously studied. Two long-term observational studies of pirfenidone are underway – RECAP and PASSPORT (Pirfenidone Post-Authorization Safety Registry). Interim results of both indicate that pirfenidone is well tolerated in the long term, with a median treatment time in the RECAP study of 163.3 weeks. Additionally, multiple single-center, 'real-world' experiences have been published. These studies suggest continued reduction in the rate of decline in lung function and reasonable tolerability. Long-term data on the safety and efficacy of nintedanib are thus far lacking. In the context of a deadly disease, it is our recommendation to continue patients on OAFs for life or until transplantation.
Both drugs have been shown to work in well established and appropriately diagnosed IPF and this is where their use should remain pending studies of these agents in other fibrotic lung disorders.
Defining successful treatment is particularly difficult when the goal of treatment is simply to slow the progression of disease. In a patient with progressive decline, the argument can always be made that perhaps their decline would have been more precipitous without therapy. Also the course of IPF is unpredictable – patients with long-term stable pulmonary function can suddenly develop progressive disease. A decline in the FVC of 10% has consistently been shown to be associated with poorer outcomes and is widely regarded as the threshold that defines 'treatment failure'. Should therapy in these patients be terminated or changed? Emerging data from an analysis of the combined CAPACITY and ASCEND cohorts of patients who, during the course of their respective studies, had a 10% FVC decline in the first 6 months on pirfenidone or placebo, demonstrated a markedly better outcome for those who continued on pirfenidone versus those in the placebo arm. The message from this suggests that patients be continued on drug no matter how significant a decline in the FVC. Whether the same holds true for nintedanab is uncertain. Another option in the face of declining lung function or rapidly deteriorating status is to change OAFs. This might not be unreasonable, although the effectiveness of this practice remains to be determined. It is certainly not unreasonable to switch OAFs if patients have intolerable side effects from one of the agents.
It is theoretically attractive to combine OAFs in an effort to target the multiple pathways leading to fibrosis in IPF. For now, there are no data demonstrating improved efficacy with combination therapy. A phase 2 Japanese study has examined the tolerability of adding nintedanib to pirfenidone in 21 patients with IPF. Ten of the 21 patients on combination therapy experienced mild or moderate adverse events, primarily gastrointestinal disturbance. While this small study lends credence to the safety of combined therapy, there are still many unknowns including the possibility of pharmacokinetic interactions between the two drugs. Lack of data has not hindered the clinical application of combination therapy in many other diseases. A prime example is the field of pulmonary arterial hypertension, where combination therapy has been used for many years and it is only recently that clinical trial evidence is emerging validating the added efficacy of such an approach. While we do not endorse combination therapy, given the lack of data, potential for side effects and expense, it might not be an unreasonable consideration in select patients, provided the appropriate discussion is held and the patient enters this relatively uncharted territory fully informed.
Commonly Asked Questions About the New Oral Antifibrotics
Which Oral Antifibrotic Should I Use?
On the basis of the available RCT data, clinical superiority of one agent over the other has not been established. Both agents have been demonstrated in multiple phase 3 trials to slow the decline in FVC in patients with mild to moderate IPF with acceptable safety profiles. It could be argued that pirfenidone holds the advantage of demonstrating a reduction in all-cause and IPF-related mortality on pooled analysis of its three major trials. On the contrary, nintedanib appears to have an effect in preventing acute exacerbations of IPF, based on central adjudication of events from the INPULSIS pooled patient populations. A recent meta-analysis performed by Loveman et al. indirectly compared the rate change in FVC and reported nintedanib had 'superior benefit' on the rate of change in FVC. It is ill-advised comparing one trial to the next to determine better efficacy of these agents. The ASCEND, CAPACITY, and INPULSIS studies were different in a number of ways. Although all the studies were in well documented IPF, nuanced inclusionary and exclusionary criteria differences resulted in slightly different populations being studied (see Table 4 ). While the primary outcome measure for all studies was the change in FVC, this change was analyzed differently for pirfenidone and nintedanib. Specifically, it was the slope generated from all FVC measures through the course of the study that was used to demonstrate the efficacy of nintedanib, whereas pirfenidone's efficacy was established through a rank analysis of covariance (ANCOVA) analysis. Both the ASCEND and INPULSIS studies were 52 weeks in duration, which seemingly would allow a comparison of change in the FVC between all groups over this unifying period. However, once again, this assumption is fraught with danger and misinterpretation. For example, patients who died in the ASCEND study were imputed a zero value for their FVCs which drove down the rate of change for both the placebo and treatment arms in this study.
Since the efficacy cannot and should not be compared, the decision as to which agent to choose is predicated to a large extent by the potential side effects and patient preference. This is based on cognisance of the patients' prior medical history, their lifestyle, and their 'comfort' zone. Indeed, a salient goal is to maintain patients on therapy since drug discontinuation is more likely to be deleterious to their subsequent outcomes than the choice of agent. Perhaps, as the underlying pathogenesis of IPF becomes more apparent, we will be able to identify likely 'responders' to one agent or another. For now, we assume clinical equipoise and base our decision on the tolerability of the individual medications.
Who Should I Start Oral Antifibrotics in? Can Patients Be Too Severe or Not Severe Enough?
There is no universally recognized system whereby patients' severity of disease is graded. Clinicians typically look at the FVC% predicted and the percent predicted diffusing capacity of the lung for carbon monoxide (DLCO%) predicted (to a lesser extent) to provide a gestalt of this. In terms of the ASCEND and INPULSIS studies, patients with 'mild to moderate' disease were enrolled as defined by the inclusion criteria shown in Table 4 . When the US FDA approved both drugs, their respective prescribing information states that they are approved 'for the treatment of IPF' without any caveat as to the severity. On the one end of the philosophical spectrum, a purist approach would be to limit prescription of these drugs to those IPF patients who meet the pulmonary function test criteria of the respective studies. Arguably then, perhaps patients would need to meet all the inclusion and exclusion criteria of the respective studies. Unfortunately, it is a small minority of IPF patients (~27%) who meet all criteria for enrollment in IPF clinical trials. A more pragmatic approach encompasses the assumption that since the drugs have been shown to work in a select IPF population, they are likely to work in other IPF patients. Given the poor prognosis of IPF and lack of alternative therapies, our practice is to consider and discuss the possibility of therapy in all IPF patients.
On the one end of this spectrum are those patients in whom IPF was diagnosed through an incidental finding and who are asymptomatic with normal lung function. It conceptually makes sense to introduce therapy as early as possible in an effort to preserve pulmonary function for as long as possible. While it might not be unreasonable to observe these patients, we believe that it is nonetheless a discussion to be had with each patient so that they may participate in this decision. Indeed, an asymptomatic patient may present months or years later with symptoms and significant loss of lung function. Since both agents slow the rate of loss of lung function and neither restores lung function, in the absence of a prior documented discussion about therapy, the practitioner might be at medicolegal jeopardy for having withheld medication that might have prevented or slowed the onset of symptoms.
What of those patients with more severe disease, specifically a FVC below 50% predicted? This population of patients are frequently excluded from clinical trials, and it is therefore unknown if oral antifibrotics (OAFs) retain the same efficacy in severe disease as in mild to moderate disease. Limited retrospective data suggest that pirfenidone slows the rate of decline in patients with FVC below 60% predicted. Furthermore, a subgroup analysis from the combined CAPACITY and ASCEND datasets did not show a difference in drug efficacy between patients above and below the median FVC% of the study population. It might be inferred from this that the severity of disease at baseline does not make a difference with regards to response, although, of course, this analysis only included patients from within these studies. Patients with an FVC below 50% of that predicted are not a homogenous group in terms of disease severity. There are those who might be New York Heart Association functional class 2 or even 3 on low flow supplemental oxygen, whereas others are tachypneic at rest on high flow oxygen. In the former situation, therapy might be entirely reasonable, whereas in the latter cases, it might be regarded as futile. Whenever treatment with an OAF is considered and prescribed, it should only occur after a frank discussion of the expectations of therapy, and the possibility and spectrum of side effects of the drugs.
How Long Should I Continue Therapy?
The efficacy of OAFs for longer than 52–72 weeks has not been rigorously studied. Two long-term observational studies of pirfenidone are underway – RECAP and PASSPORT (Pirfenidone Post-Authorization Safety Registry). Interim results of both indicate that pirfenidone is well tolerated in the long term, with a median treatment time in the RECAP study of 163.3 weeks. Additionally, multiple single-center, 'real-world' experiences have been published. These studies suggest continued reduction in the rate of decline in lung function and reasonable tolerability. Long-term data on the safety and efficacy of nintedanib are thus far lacking. In the context of a deadly disease, it is our recommendation to continue patients on OAFs for life or until transplantation.
Should Either Drug Be Used in Other Forms of Fibrotic Lung Disease?
Both drugs have been shown to work in well established and appropriately diagnosed IPF and this is where their use should remain pending studies of these agents in other fibrotic lung disorders.
What Defines a Treatment Failure (or Success)? When (if Ever) Should I Switch Oral Antifibrotics?
Defining successful treatment is particularly difficult when the goal of treatment is simply to slow the progression of disease. In a patient with progressive decline, the argument can always be made that perhaps their decline would have been more precipitous without therapy. Also the course of IPF is unpredictable – patients with long-term stable pulmonary function can suddenly develop progressive disease. A decline in the FVC of 10% has consistently been shown to be associated with poorer outcomes and is widely regarded as the threshold that defines 'treatment failure'. Should therapy in these patients be terminated or changed? Emerging data from an analysis of the combined CAPACITY and ASCEND cohorts of patients who, during the course of their respective studies, had a 10% FVC decline in the first 6 months on pirfenidone or placebo, demonstrated a markedly better outcome for those who continued on pirfenidone versus those in the placebo arm. The message from this suggests that patients be continued on drug no matter how significant a decline in the FVC. Whether the same holds true for nintedanab is uncertain. Another option in the face of declining lung function or rapidly deteriorating status is to change OAFs. This might not be unreasonable, although the effectiveness of this practice remains to be determined. It is certainly not unreasonable to switch OAFs if patients have intolerable side effects from one of the agents.
Should I Ever Combine Pirfenidone and Nintedanib?
It is theoretically attractive to combine OAFs in an effort to target the multiple pathways leading to fibrosis in IPF. For now, there are no data demonstrating improved efficacy with combination therapy. A phase 2 Japanese study has examined the tolerability of adding nintedanib to pirfenidone in 21 patients with IPF. Ten of the 21 patients on combination therapy experienced mild or moderate adverse events, primarily gastrointestinal disturbance. While this small study lends credence to the safety of combined therapy, there are still many unknowns including the possibility of pharmacokinetic interactions between the two drugs. Lack of data has not hindered the clinical application of combination therapy in many other diseases. A prime example is the field of pulmonary arterial hypertension, where combination therapy has been used for many years and it is only recently that clinical trial evidence is emerging validating the added efficacy of such an approach. While we do not endorse combination therapy, given the lack of data, potential for side effects and expense, it might not be an unreasonable consideration in select patients, provided the appropriate discussion is held and the patient enters this relatively uncharted territory fully informed.
Source...