A cough that outlasts a job, a mortgage or a marriage is not a rare problem, and a Canadian cough clinic has now measured how well the recommended playbook actually works on it. The answer, published August 24 in ERJ Open Research, is sobering: patients coughed measurably less, and most of them still did not feel their cough was under control.
The finding lands about five weeks after the pharmaceutical industry's most advanced candidate for the condition was shelved, leaving clinicians with the same off-label toolkit they have been improvising with for a decade.
A Cough Clinic Measured What Guideline Care Really Delivers
PROCOUGH was a prospective observational cohort run at McMaster University Medical Centre in Hamilton, Ontario, between December 2021 and May 2025. One hundred adults with cough lasting longer than eight weeks completed the study. Their average age was 58, 58 percent were women, and the median cough duration was seven years.
These were not casual referrals. Every patient underwent lung function testing, airway inflammation measures, sputum cytology and 24-hour objective cough monitoring with a wearable recorder, then received treatment following European Respiratory Society guideline recommendations. Everything was measured again after the treatment trial.
Patients with an identifiable driver such as asthma or eosinophilic bronchitis received targeted therapy, most often an inhaled corticosteroid with a long-acting beta agonist. Those whose testing came back clean, or who had already failed those treatments, were offered either a neuromodulator, which in this cohort meant morphine, pregabalin, gabapentin or amitriptyline, or virtual cough control therapy delivered by a physiotherapist and a speech-language pathologist.
The Numbers Improved While the Patients Did Not Feel Better
Objectively, treatment worked. Twenty-four-hour cough frequency fell by 54 percent, with a 95 percent confidence interval of 42.0 to 62.9 percent. Quality of life on the Leicester Cough Questionnaire rose 3.4 points, comfortably past the 1.3-point threshold researchers treat as meaningful. Cough severity fell by 20.2 millimeters on a 100-millimeter scale, which is real but short of the 30 millimeter change the field uses as its own benchmark.
Then the investigators asked patients a blunt question: is your cough sufficiently better that you feel no further treatment is needed?
Seventy-one percent said no.
The responder analyses show the same split. Fifty-six percent of patients hit the objective threshold of a 30 percent or greater cut in cough frequency, 64 percent cleared the quality-of-life threshold, and only 36 percent cleared the cough severity threshold. Just 14 percent cleared all three at once. Improvement and control turned out to be different things.
That gap is easier to understand with the burden in view. Chronic cough is not a nuisance symptom in this population. The authors describe patients living with disturbed sleep, urinary incontinence, fainting, anxiety and social withdrawal, and the people in this cohort had been coughing for a median of seven years before the treatment trial began. Cutting a cough in half still leaves a lot of coughing.
The researchers concluded that guideline-based therapy improves objective and subjective outcomes but that most patients fail to reach satisfactory symptom control. Their study has real limits. It was a single center, it had no placebo comparison group, and patients referred to a tertiary cough clinic are by definition the harder cases. The authors also note that the 54 percent drop sits within the range of placebo responses seen in blinded trials of refractory cough.
The Drug Meant to Fill That Gap Was Abandoned in July
On July 17, GSK announced that it would stop developing camlipixant for refractory chronic cough after reviewing its two pivotal phase 3 trials.
CALM-1 met its primary endpoint, with camlipixant 50 mg twice daily producing a statistically significant reduction in 24-hour cough frequency against placebo at week 12. CALM-2 did not reach significance on the same measure at week 24. The 25 mg dose missed in both studies, and key secondary endpoints including a chronic cough diary measure fell short in both. Treatment-related side effects were similar to placebo.
GSK said the limited efficacy it saw was unlikely to transform patient care. The company had acquired the drug through its purchase of Bellus Health in 2023. A phase 2b trial of camlipixant in irritable bowel syndrome continues.
Camlipixant blocked P2X3, a receptor on airway sensory nerves implicated in the hypersensitive cough reflex, and had shown promise in mid-stage testing. It was the most advanced attempt to give American patients a drug licensed specifically for this condition, and there is no obvious replacement in late-stage development.
The Exotic Explanations Are Real but Rare
Unexplained cough occasionally has a startling cause. A case report published in July in BMC Pulmonary Medicine described a 61-year-old woman whose year-long cough, worse after eating and drinking, was eventually traced to cerebellar tonsillar herniation. Surgery to decompress the base of her skull ended the cough, and it had not returned a year later.
Cases like that are worth knowing about and easy to over-read. Chronic cough affects roughly one adult in ten worldwide, with regional estimates running from about 2 to 18 percent. The overwhelming majority of those coughs are not hiding a brainstem problem, a tumor or a foreign body. They are ordinary refractory coughs in ordinary people whose testing comes back clean.
For those patients, the practical picture after this summer is unchanged. Guideline care helps, speech and cough control therapy helps, and neither reliably ends the problem. Anyone coughing for more than eight weeks should be evaluated rather than self-treating, particularly if there is blood, weight loss, fever, or breathlessness. Anyone already through that workup should know that partial relief is a common outcome, not a personal failure.
Key Questions Answered
What counts as chronic cough?
A cough lasting longer than eight weeks. It is called refractory when it persists despite treatment of identifiable causes, and unexplained when testing finds no cause at all.
What did the PROCOUGH study find?
In 100 patients treated by guideline recommendations, 24-hour cough frequency fell 54 percent, but 71 percent still reported their cough was not controlled well enough to stop treatment.
Does this mean treatment is pointless?
No. Objective and patient-reported measures both improved significantly. The study shows that meaningful improvement often falls short of the relief patients actually want.
Why was camlipixant stopped?
One phase 3 trial met its primary endpoint and the other did not, and key secondary endpoints were missed in both. GSK judged the overall efficacy unlikely to transform patient care.
Is there an approved drug for this in the United States?
No drug is licensed in the US specifically for refractory or unexplained chronic cough. Clinicians rely on off-label treatments and non-drug cough control therapy.
When should a persistent cough be investigated urgently?
Coughing up blood, unexplained weight loss, fever, night sweats, breathlessness or a change in an existing cough all warrant prompt medical assessment.
