Pharmacological Interventions for Breathlessness—Absence of Effect or Absence of Evidence?

Breathlessness is a common, highly distressing symptom for people with advanced cancer and their caregivers. Breathlessness may be acute or chronic and, for 70% of patients, is episodic, most often triggered by exertion.¹ Importantly, breathlessness in advanced cancer is a challenging symptom for clinicians to manage because the underlying physiological mechanisms are diverse and complex, ranging from complications such as airway obstruction, pulmonary metastases, lymphangitis, radiation pneumonitis and fibrosis, pleural effusions, pneumonia, anemia, pulmonary emboli, myopathy, cachexia, and others. Additionally, comorbidities that share common risk factors with cancer (eg, chronic obstructive pulmonary disease) may also contribute. These diverse conditions lead to afferent receptor stimulation and increased respiratory neural drive from the respiratory center in the brainstem. Central perception of breathlessness arises through the complex interplay of higher cortical awareness of the mismatch between respiratory neural activity drive and mechanical and ventilatory efforts as well as psychosocial and cultural factors, which influence individual cognitive interpretation of the symptom.²

Given the many physiological mechanisms that generate breathlessness, a comprehensive management approach must be multimodal, aiming to first treat the underlying cancer and all reversible causes, with nonpharmacological and pharmacological interventions added to address persisting breathlessness. Opioids have been the most studied pharmacological agents because there are numerous opioid receptors both centrally and peripherally, with postulated mechanisms of action including reduced respiratory drive and blunting of perceptual sensitivity to breathlessness.

Elsewhere in JAMA Network Open, Feliciano et al³ report findings from their systematic review, which analyzed data from 19 studies (17 randomized clinical trials and 2 retrospective observational studies with 1424 patients), regarding the benefits and harms of medications (mainly opioids and anxiolytics) for breathlessness in people with advanced cancer. The authors found that pharmacological agents were not associated with more effectiveness than placebo for the treatment of breathlessness, anxiety, or exercise capacity. The data were inconclusive regarding any changes in quality of life. Dose and route of administration of opioids did not alter the findings, and pharmacological interventions were associated with more harms than placebo (albeit infrequently reported). The authors rated much of the evidence as low quality.³

Analyzing the evidence for pharmacological interventions for breathlessness through meta-analyses is complex and challenging. Feliciano et al³ defined advanced cancer as cancer unlikely to be cured with treatment and rightly noted that this phase of illness is no longer synonymous with end of life. Instead, people with advanced cancer represent a diverse group; the evolution of new cancer therapies means some people live with advanced disease for many years. Additionally, while most patients had lung cancer or mesothelioma, the studies within the systematic review included people with nearly all types of cancer. Thus, the population treated was highly heterogeneous, with different prognoses, and recruited from multiple different settings (ie, inpatients, outpatients, and those receiving palliative care).

Most trials included in the systematic review were small, short-term studies, which measured breathlessness at different points (ranging from 6-172 minutes) after drug administration, either at rest, peak exercise, or isotime. Only 7 studies evaluated effects beyond 24 hours, with 2 studies examining effects at 2 weeks and 1 at 4 weeks. Most studies focused on opioids; however, different doses, formulations, and routes of administration were used. The included studies also used different unidimensional scales to assess breathlessness. As the authors outline,³ there is much debate in the literature regarding the optimal tool to measure dyspnea. As breathlessness has multiple dimensions, including intensity, unpleasantness, associated emotional response, and functional impact,⁴ it is now generally accepted that breathlessness should be measured on a multidimensional scale, which none of the included studies used. Nevertheless, the specific multidimensional scale used, at what time, and during what activities remains unclear.

Combining such heterogeneous factors requires the use of a standardized mean difference, which may underestimate the true effect. Some included studies were crossover trials, which may be analyzed as parallel studies at the cost of a wider confidence interval or require sufficient reported trial information to impute a corrected standard error. These studies also vary on whether they report change from baseline or posttreatment scores. Mean differences based on changes from baseline may be assumed to be the same as posttreatment scores, although this can come at the cost of precision. However, if posttreatment scores and change from baseline scores are to be combined, standardized mean differences are not recommended (because the standard deviations for the change scores are not the same as the standard deviations for the final scores).⁵ The requirement to transform the data multiple times leads to difficulty in the interpretation of the overall result.

Given the heterogeneity in patient populations and settings, different study designs, drugs, measurements, statistical analyses, and lack of data on harms, Feliciano et al³ must be commended for their excellent review. Their findings are not dissimilar from the findings from the 2016 Cochrane review on this topic, which included 26 studies with 526 participants with either advanced malignant or nonmalignant conditions.² In that study, Barnes et al² concluded that there was only some low-quality evidence, which showed a small benefit for the use of oral or parenteral opioids for breathlessness. The authors² identified a need for further research with larger studies, using standardized protocols and with quality-of-life measures included.

Nevertheless, it is clear from the systematic review by Feliciano et al³ that the evidence remains poor. Indeed, the authors’ conclusion must be interpreted with caution, given that the absence of evidence is not absence of effect. The truth is that still we do not know whether there is no benefit from pharmacological treatments for breathlessness in people with advanced cancer or if the findings from Feliciano et al³ are due to the limited, low-quality evidence in highly heterogeneous populations. Therefore, we suggest that the main message from the study by Feliciano et al³ is a call to arms: we urgently need high-quality, large, long-term, multisite randomized clinical trials that robustly examine and report both benefits (particularly regarding quality of life) and harms of pharmacological interventions for breathlessness over multiple points in more homogeneous groups of people with advanced cancer. Furthermore, the risk of harms must be considered in the context of survival and the use of opioids for breathlessness for many years. These longer-term survivors are likely to be independent individuals who drive, yet no study to date has reported the effects of opioids on driving when prescribed to manage breathlessness.

While clinical trials at the very end of life pose many ethical risks and challenges, it is clear that many people with advanced care are not in the terminal phase, so they could be suitable for inclusion in high-quality studies to address this question. Nevertheless, the challenges of obtaining funding, gaining ethical approval, recruiting participants, and running expensive clinical trials are real and in part explain why many studies to date have been small, short term, and rated low quality. While curing cancer remains the ultimate goal, our patients come to us with symptoms for which we have no good evidence-based treatments. Clinical research to address challenging symptoms, such as distressing breathlessness, remains urgently needed so that we can find novel, effective treatments for our patients.

Published: February 25, 2021. doi:10.1001/jamanetworkopen.2021.0262

Open Access: This is an open access article distributed under the terms of the CC-BY License. © 2021 Barnes H et al. JAMA Network Open.

Corresponding Author: Hayley Barnes, MBBS, MPH, Central Clinical School, Monash University, Level 5, 99 Commercial Rd, Melbourne 3004, VIC Australia (hayley.barnes@monash.edu).

Conflict of Interest Disclosures: Dr Barnes reported receiving support from the National Health and Medical Research Council through a Postgraduate Scholarship during the conduct of the study. No other disclosures were reported.