Effect of dexamethasone on dyspnoea in patients with cancer (ABCD): a parallel-group, double-blind, randomised, controlled trial

Lancet Oncol. 2022 Oct;23(10):1321-1331. doi: 10.1016/S1470-2045(22)00508-3. Epub 2022 Sep 7.


Background: Systemic corticosteroids are commonly prescribed for palliation of dyspnoea in patients with cancer, despite scarce evidence to support their use. We aimed to assess the effect of high-dose dexamethasone versus placebo on cancer-related dyspnoea.

Methods: The parallel-group, double-blind, randomised, controlled ABCD (Alleviating Breathlessness in Cancer Patients with Dexamethasone) trial was done at the at the University of Texas MD Anderson Cancer Center and the general oncology clinic at Lyndon B Johnson General Hospital (both in Houston, TX, USA). Ambulatory patients with cancer, aged 18 years or older, and with an average dyspnoea intensity score on an 11-point numerical rating scale (NRS; 0=none, 10=worst) over the past week of 4 or higher were randomly assigned (2:1) to receive dexamethasone 8 mg orally every 12 h for 7 days followed by 4 mg orally every 12 h for 7 days, or matching placebo capsules for 14 days. Pharmacists did permuted block randomisation with a block size of six, and patients were stratified by baseline dyspnoea score (4-6 vs 7-10) and study site. Patients, research staff, and clinicians were masked to group assignment. The primary outcome was change in dyspnoea NRS intensity over the past 24 h from baseline to day 7 (±2 days). Analyses were done by modified intention-to-treat (ie, including all patients who were randomly assigned and started the study treatment, regardless of whether they completed the study). Enrolment was stopped after the second preplanned interim analysis, when the futility criterion was met. This study is registered with ClinicalTrials.gov (NCT03367156) and is now completed.

Findings: Between Jan 11, 2018, and April 23, 2021, we screened 2867 patients, enrolled 149 patients, and randomly assigned 128 to dexamethasone (n=85) or placebo (n=43). The mean change in dyspnoea NRS intensity from baseline to day 7 (±2 days) was -1·6 (95% CI -2·0 to -1·2) in the dexamethasone group and -1·6 (-2·3 to -0·9) in the placebo group, with no significant between-group difference (mean 0 [95% CI -0·8 to 0·7]; p=0·48). The most common all-cause grade 3-4 adverse events were infections (nine [11%] of 85 patients in the dexamethasone group vs three [7%] of 43 in the placebo group), insomnia (seven [8%] vs one [2%]), and neuropsychiatric symptoms (three [4%] vs none [0%]). Serious adverse events, all resulting in hospital admissions, were reported in 24 (28%) of 85 patients in the dexamethasone group and in three (7%) of 43 patients in the placebo group. No treatment-related deaths occurred in either group.

Interpretation: High-dose dexamethasone did not improve dyspnoea in patients with cancer more effectively than placebo and was associated with a higher frequency of adverse events. These data suggest that dexamethasone should not be routinely given to unselected patients with cancer for palliation of dyspnoea.

Funding: US National Cancer Institute.

Publication types

  • Randomized Controlled Trial
  • Research Support, N.I.H., Extramural

MeSH terms

  • Adrenal Cortex Hormones / therapeutic use
  • Dexamethasone / adverse effects
  • Double-Blind Method
  • Dyspnea / chemically induced
  • Dyspnea / etiology
  • Humans
  • Neoplasms* / drug therapy
  • Treatment Outcome


  • Adrenal Cortex Hormones
  • Dexamethasone

Associated data

  • ClinicalTrials.gov/NCT03367156