Breathlessness, dyspnoea and shortness of breath name the same symptom. The European Society for Medical Oncology (ESMO), in its 2020 guideline, defines it as a subjective experience of breathing discomfort that varies in intensity, and notes that people with cancer often cite it as their most distressing symptom. The trials below score it on a 0 to 10 numerical rating scale, a 10 cm visual analogue scale (VAS), or the modified Borg scale, a 0 to 10 rating of how hard breathing feels. Most were run in lung cancer and mesothelioma; this page also reports what the reviews found in heart failure and other advanced disease. Trials in chronic obstructive pulmonary disease are in COPD.
No current Cochrane review
A 2008 Cochrane review by Bausewein and colleagues of non-drug treatments for breathlessness in advanced cancer and other advanced disease found 5 studies of acupuncture or acupressure and rated the evidence that they help as low in strength. Cochrane withdrew the review in 2013 as out of date. COPD sets out what happened to its planned replacement, and why the evidence on this page comes from reviews in other journals.
Reviews
A standardised mean difference (SMD), or Hedges’ g, states an effect in units of the spread between patients, so that trials using different scales can be pooled; a negative value here means less breathlessness. A network meta-analysis compares many treatments at once, including pairs no trial tested head to head, and ranks them. The four reviews below differ in what they compared acupuncture or acupressure with, and the table keeps that apart.
| Review | Trials and people | Comparison | Finding on breathlessness | Certainty |
|---|---|---|---|---|
| von Trott 2020, acupuncture and acupressure in advanced disease | 12 trials, 597 people (347 with COPD, 190 with advanced cancer) | Any control; a subgroup against sham | SMD −1.77 (95% CI −3.05 to −0.49; 10 trials, 480 people; I² 90%) | Not graded in the abstract |
| Gupta 2021, all non-drug treatments in advanced cancer, searched to May 2020 | 29 trials, 2,423 people; 4 of integrative medicine | Any control | Acupressure and reflexology associated with less breathlessness in outpatients | Low strength of evidence |
| Vo 2025, network meta-analysis in advanced cancer, searched to March 2024 | 42 trials, 3,832 people | Control groups and drug and non-drug treatments | Acupressure or reflexology SMD −1.04 (95% CI −2.02 to −0.06); acupuncture ranked last of 17 treatments | Very low (GRADE) for acupressure or reflexology |
| Hu 2026, network meta-analysis in lung cancer, searched to June 2025 | 46 trials, 4,138 people; 2 of acupressure, 2 of acupuncture | Usual care | Acupressure Hedges’ g −1.96 (95% CI −2.92 to −0.99), ranked first of 10 strategies; acupuncture no significant effect | Moderate for acupressure against usual care; low for most comparisons (CINeMA) |
Von Trott and colleagues, at King’s College London’s Cicely Saunders Institute and Havelhoehe Hospital in Berlin, included trials in which at least a quarter of participants had advanced disease. In 6 trials with 302 people that used a sham control and treated for at least 3 weeks, the difference was larger (SMD −2.53, I² 91%); I² is the share of the variation between trials that chance does not explain. The authors named the varied methods and low power of the trials as problems, and asked future trials to test whether acupuncture added to drugs reduces the need for morphine.
Gupta and colleagues, at Johns Hopkins, screened only English-language studies. Vo and colleagues, in Ho Chi Minh City and Taipei, pooled acupressure with reflexology, foot pressure on reflex zones, as one treatment, so their estimate does not isolate acupressure. Hu and colleagues, at Sun Yat-sen University and Johns Hopkins, wrote that only two acupressure trials were included, one with four participants over 3 days, and that their ranking should be read with caution. Their two acupuncture trials, from Korea and China, enrolled people with early or locally advanced lung cancer.
For heart failure, a 2025 review by Ganglani and Geng in Houston included 30 studies and combined their p values rather than their effect sizes. The authors wrote that heterogeneity between the studies limited any generalisation and that a pilot trial was still needed.
Trials of acupuncture in cancer
In a 1996 open pilot at the Royal Marsden Hospital in London, Filshie and colleagues needled points on the upper sternum and LI04 (Hegu) in 20 people breathless at rest because of cancer. 14 of the 20 reported marked benefit, and breathlessness scores fell for at least 6 hours, most at 90 minutes. There was no control group.
A 2005 sham-controlled pilot by Andrew Vickers and colleagues at Memorial Sloan-Kettering Cancer Center in New York randomised 47 people with lung or breast cancer to one session of true or placebo acupuncture on top of their usual treatment, followed by semi-permanent studs that they pressed twice a day for a week. True acupuncture used LU01 (Zhongfu), LU07 (Lieque), ST36 (Zusanli), KD06 (Zhaohai), CV06 (Qihai), two ear points and two needles into the upper sternum; the placebo used points away from the true ones. Breathlessness scores were 0.34 points higher with true acupuncture straight after treatment (95% CI −0.33 to 1.02) and 0.56 higher over the following week (−0.39 to 1.51). The authors concluded that the technique was “unlikely to have effects on dyspnea importantly larger than placebo”. No adverse event was attributed to acupuncture.
The Royal Marsden group then compared acupuncture with morphine. In a 2016 phase II trial, Minchom and colleagues randomised 173 people with non-small cell lung cancer or mesothelioma and a breathlessness score of 4 or more on the VAS to acupuncture alone, morphine alone, or both. Acupuncture used points on the upper sternum, beside the thoracic spine, in the trapezius muscle and at LI4, with studs left in the top of the sternum for patients to massage when breathless; the acupuncture-alone arm could take morphine as rescue. A fall of at least 1.5 points at 4 hours, the primary end point, was reached by 74% with acupuncture, 60% with morphine and 66% with both, with no significant difference between the arms. At least one dose of morphine was taken by 21% of the acupuncture arm and 87% of each of the other two. The trial had no sham and patients knew their treatment.
Both trials put needles over the chest, into or beside the sternum and, in Vickers’ trial, at LU01 in the first intercostal space. Needling over the chest reports what the case series record about needles reaching the lung, and Acupuncture in cancer care covers the other precautions trials apply in people with cancer.
In a 2016 feasibility study at the University of Pennsylvania, Bauml and colleagues gave 12 people with lung cancer 10 weekly sessions, with no control group. Breathlessness on the numerical rating scale fell from 6.3 to 3.6 at the end of treatment and was 3.2 four weeks later; 10 of the 12 completed all sessions, and adverse events were mild.
Trials of acupressure
Two trials in Turkey taught or applied pressure at the same three points, PC06 (Neiguan), LU01 and LU10 (Yuji), against standard care with no sham. In a 2020 trial, Doğan and Taşcı gave acupressure twice a day for 4 weeks to people with lung cancer having chemotherapy, and reported lower breathlessness and better quality of life than in the standard care group. In a 2025 trial, Kocatepe and colleagues randomised 115 people in a palliative care unit, 57 to acupressure for 3 minutes twice a day for 14 days and 58 to standard care. On day 14 the modified Borg score was 3.0 with acupressure and 5.4 with standard care.
What guidelines say
The ESMO guideline, published in ESMO Open in December 2020, records its last update as October 2020 and supersedes the 2015 version. It recommends considering a therapeutic trial of acupressure or acupuncture according to patient preference, graded II, C: level II means small randomised trials, or larger ones or meta-analyses with a suspicion of bias or heterogeneity, and grade C means insufficient evidence for efficacy, so the recommendation is optional. It cites von Trott’s review and Minchom’s trial, and asks for high-quality studies of the long-term risks and benefits.
The American Society of Clinical Oncology (ASCO) guideline by Hui and colleagues, published in the Journal of Clinical Oncology in February 2021, drew on a review of 48 randomised trials and two cohort studies by the Agency for Healthcare Research and Quality. It states that psychoeducational, self-management or complementary approaches may be offered, and recommends systemic opioids when non-drug measures do not give enough relief; its PubMed record lists no update. The ASCO Post report on the guideline names acupressure or reflexology among the interventions that may be offered; it does not name acupuncture. Clinical guidelines compares other bodies’ positions on acupuncture.
What the reviews could not settle
The one sham-controlled acupuncture trial in cancer, Vickers’, gave a single session to 47 people and found no difference from placebo, while von Trott’s sham-controlled subgroup, mostly in COPD, found a large one with an I² of 91%. Hu’s and Vo’s network analyses ranked acupressure first or second, but each rests on a few trials against usual care with no sham, and Vo’s merges acupressure with reflexology. Minchom’s comparison with morphine had no sham and no blinding. Outside cancer and COPD, the one review found, Ganglani’s in heart failure, combined p values rather than effects, so it gives no estimate of how large any difference is, and none of the reviews reports a separate estimate for interstitial lung disease.
AcuiQ’s dyspnoea and shortness of breath pages list the protocols individual studies prescribed, each with its citation. Cancer-related fatigue reports the trials for fatigue, which Bauml’s study also measured.