The mechanistic case against opioids is, on its own terms, coherent and reproducible. Mu-opioid receptors, the molecular docking sites targeted by morphine and related drugs, are expressed not only on neurons but also on key cells of the immune system, including natural killer cells and CD8-positive T lymphocytes, the two cell types most responsible for recognizing and destroying tumor cells. In vitro experiments and animal models have repeatedly shown that activating these receptors can dampen the cytotoxic activity of NK cells, impair T-cell proliferation and function, and dysregulate the cytokine signaling networks that coordinate an immune response. Opioids can also act centrally, engaging the hypothalamic-pituitary-adrenal axis and the sympathoadrenal system to raise levels of cortisol and catecholamines, stress hormones that are themselves immunosuppressive. In animal studies, these effects have sometimes translated into faster tumor growth or greater metastatic spread.

The problem, Mercadante argues, is the gap between the petri dish and the clinic. When researchers move from controlled animal experiments to human studies, the evidence base changes character entirely. Clinical data on opioids and cancer outcomes come almost exclusively from retrospective cohort analyses, in which investigators look back at existing patient records rather than randomly assigning treatment. Such studies are notoriously vulnerable to a statistical trap known as confounding by indication. Patients who require strong opioids are, almost by definition, those with more advanced disease, larger tumor burdens, worse performance status, and greater systemic inflammation. Each of those factors independently predicts poorer survival and weaker responses to immunotherapy, regardless of what painkillers the patient happens to receive. The opioid, in other words, may simply be a marker of how sick someone is, not a cause of it.

That confounding problem becomes even sharper when the outcome in question is response to immune checkpoint inhibitors, the antibody drugs that unleash T cells against tumors by blocking inhibitory receptors such as PD-1 and PD-L1. Several retrospective analyses and at least one systematic review and meta-analysis have reported that cancer patients taking opioids fare worse on checkpoint inhibitors than those who are not. But the same logic applies: patients needing opioids for breakthrough pain are more likely to have bulky, symptomatic, late-stage disease, and it is the disease, not necessarily the drug, that drives the inferior outcomes. Without pain-matched comparators, patients with equally advanced cancer but comparable pain control achieved without opioids, the analyses cannot disentangle pharmacology from disease severity.

Mercadante also spotlights a counter-argument that he describes as largely neglected in the literature: uncontrolled cancer pain is itself a potent immunosuppressant. Severe pain activates the same HPA-axis and sympathoadrenal stress pathways that opioids can engage, flooding the body with cortisol and catecholamines that suppress NK cell activity and T-cell function. Pain is not merely a symptom to be tolerated for the sake of immune health; it is an active physiological insult. Systematic reviews have shown that pain severity independently predicts shorter survival in cancer patients, and recent work in older adults with newly diagnosed cancer has linked patient-reported pain severity to reduced survival. In this light, inferior outcomes among opioid users may reflect the disease burden and pain intensity that necessitated opioid therapy in the first place, rather than a direct pharmacological effect of the drugs.

The clinical literature on opioids and infection risk tells a similarly ambiguous story. Some observational studies in patients with cancer pain have examined whether morphine or oxycodone administration correlates with the development of infections, but the findings are inconsistent and suffer from the same confounding by indication. Immunocompromised, advanced-stage patients are more likely to need high-dose opioids and are also more likely to develop infections, making it impossible to attribute causality from retrospective data. Meanwhile, laboratory work continues to show that different opioids do not share identical immunomodulatory properties; some agents appear more immunosuppressive than others in vitro, and genetic factors may modify individual responses. These nuances, while scientifically interesting, have not yet produced a clinically actionable rule for prescribing.

One intriguing line of evidence has come from studies of methylnaltrexone, a peripherally acting mu-opioid receptor antagonist used to treat opioid-induced constipation. A retrospective analysis of patients with advanced cancer suggested that treatment with methylnaltrexone was associated with increased survival, raising speculation that blocking peripheral opioid receptors might unleash antitumor immunity. But this observation, too, comes from uncontrolled data, and alternative explanations, including healthier patients tolerating the drug better, have not been excluded. Network-based computational analyses have also suggested that opioids and checkpoint inhibitors differentially regulate overlapping immune signaling pathways in tumors such as triple-negative breast cancer, providing a plausible molecular framework for interaction but not proof of clinical harm.

Weighing the totality of the evidence, the review concludes that the preclinical signal of opioid immunosuppression, while real and mechanistically coherent, has not been validated in clinical studies of sufficient quality to justify withholding or limiting opioid analgesia in cancer patients. This includes patients receiving immune checkpoint inhibitors, a population in whom clinicians might be especially tempted to minimize opioid exposure out of fear of compromising immunotherapy. Current guidelines from bodies such as the National Comprehensive Cancer Network and the American Society of Clinical Oncology already establish effective pain management as a clinical priority in cancer care, and the new analysis reinforces that stance: the known harms of unrelieved pain, physiological, psychological, and prognostic, far outweigh a theoretical immunological risk supported only by confounded retrospective data.

The path forward, according to the review, is not practice change but better research. Prospective studies designed with pain-matched comparators are needed, in which patients with equivalent cancer stages and equivalent pain intensity are compared across different analgesic strategies, including opioid-free regimens where feasible. Such designs would finally allow investigators to separate the pharmacological effects of opioids from the effects of the disease and the pain they treat. Until those studies are done, the review’s message to oncologists, palliative care specialists, and patients is straightforward: cancer pain should be treated effectively, opioids should not be withheld out of immunological fear, and the question of whether these drugs subtly shape tumor immunity remains an open scientific question rather than a settled clinical danger.

The broader significance of the analysis lies in its methodological honesty. In an era when immunotherapy has transformed outcomes in melanoma, lung cancer, and many other malignancies, every drug a patient takes is now scrutinized for possible immune interactions, and opioids, taken by millions of people with cancer worldwide, are a natural target of that scrutiny. But the review serves as a reminder that association is not causation, that the sickest patients receive the most treatments, and that a laboratory mechanism demonstrated in mice and cell cultures does not automatically translate into a bedside mandate. For now, the evidence says the most immunosuppressive thing a clinician can do for a patient in severe cancer pain is to leave that pain untreated.

Subject of Research: The immunosuppressive effects of opioids in cancer pain management and their potential interaction with antitumor immunity and immunotherapy

Article Title: Opioids and immunosuppression in patients with cancer

Article References: Mercadante, S. (2026). Opioids and immunosuppression in patients with cancer. Supportive Care in Cancer, 34(10), Article 1031. https://doi.org/10.1007/s00520-026-11279-5

Image Credits: AI Generated

DOI: 10.1007/s00520-026-11279-5

Keywords: opioids, immunosuppression, cancer pain, morphine, immune checkpoint inhibitors, natural killer cells, T cells, confounding by indication, HPA axis, immunotherapy, palliative care, Supportive Care in Cancer