| Non-Thermal Plasma (NTP; Cold Atmospheric Plasma, CAP) — Partially ionized gas generated under conditions in which the bulk gas remains near physiological or ambient temperature. Biomedical non-thermal plasma produces a complex mixture of reactive oxygen and nitrogen species (RONS), charged particles, electric fields, photons, and other reactive components capable of modifying cellular redox state, membranes, proteins, DNA, mitochondria, and intracellular signaling.
Product / intervention class: Physical / redox-based therapeutic intervention. Cold atmospheric plasma (CAP) is the most common biomedical form of non-thermal plasma and is being investigated as an emerging anticancer treatment modality rather than as conventional chemotherapy.
Cancer relevance: Strong preclinical evidence with early clinical translation. NTP/CAP has demonstrated selective cytotoxicity against numerous cancer types, including glioblastoma, melanoma, breast, colorectal, prostate, lung, ovarian, pancreatic, and other tumors. Biological response depends strongly on plasma source, feed gas, exposure parameters, distance, treatment duration, cancer phenotype, and whether plasma is applied directly or indirectly through plasma-activated liquids.
Primary anticancer mechanism: Reactive oxygen and nitrogen species ↑ → oxidative/redox stress ↑ → mitochondrial dysfunction ↑ → apoptosis, ferroptosis, pyroptosis, or other regulated cell-death pathways ↑. NTP can also induce immunogenic cell death, suppress proliferation, inhibit EMT/migration, alter oncogenic signaling, and modulate the tumor immune microenvironment.
Typical target/pathway effects: ROS ↑; RNS ↑; mitochondrial ROS ↑; oxidative stress ↑; mitochondrial membrane dysfunction ↑; apoptosis ↑; caspase activation ↑; ferroptosis ↑ in susceptible models; lipid peroxidation ↑; DNA damage ↑; proliferation ↓; migration ↓; EMT ↓; immunogenic cell death ↑; tumor growth ↓. Exact pathway direction depends on plasma parameters and cancer type.
Clinical evidence: Early-stage clinical studies have evaluated CAP in oncology, including intraoperative treatment of residual tumor margins. Current evidence remains much stronger in vitro and in animal models than in controlled clinical trials, and standardized treatment dosimetry remains an important limitation.
Important treatment variables: Plasma source/device; dielectric-barrier discharge versus plasma jet; feed gas such as helium, argon, or air; voltage and waveform; frequency; pulse width; gas flow; treatment distance; exposure duration; delivered RONS; direct versus indirect exposure; and plasma-activated medium composition.
Non-Thermal Plasma / Cold Atmospheric Plasma — Non-thermal plasma (NTP), most commonly used biomedically as cold atmospheric plasma (CAP), is a partially ionized gas generated at or near atmospheric pressure in which electrons are energetic while the bulk gas remains near ambient or physiological temperature. CAP is a physical/redox therapeutic modality rather than a drug and produces a variable mixture of reactive oxygen and nitrogen species (RONS), charged particles, transient electric fields, excited species and photons. Standard abbreviations are NTP and CAP; related terminology includes cold atmospheric-pressure plasma and non-thermal atmospheric-pressure plasma. Biomedical CAP is generally generated using plasma jets, dielectric-barrier discharges or related electrical discharge systems using air, helium, argon or other feed gases. Direct CAP exposes tissue to the plasma plume, whereas indirect treatment uses plasma-activated liquids or media containing longer-lived reactive products. Anticancer activity is highly device- and dose-dependent, so CAP should be regarded as a treatment platform rather than a single chemically defined agent.
Primary mechanisms (ranked):
- RONS-mediated oxidative stress — extracellular and intracellular ROS/RNS accumulation can exceed the relatively limited redox reserve of susceptible cancer cells, producing selective or partially selective cytotoxicity.
- Mitochondrial dysfunction and regulated cell death — mitochondrial ROS, membrane-potential disruption, cytochrome-c release and caspase activation promote apoptosis; ferroptosis, pyroptosis and other death programs occur in specific models.
- Immunogenic cell death — CAP can increase calreticulin exposure, ATP release and HMGB1 release and thereby enhance tumor immunogenicity and antitumor immune responses.
- DNA oxidative damage and cell-cycle disruption — oxidative base modification, DNA strand breaks and checkpoint activation can produce G1 or G2/M arrest, senescence and apoptosis.
- Suppression of PI3K/AKT/mTOR/HIF-1α and glycolytic survival signaling — demonstrated particularly in hepatocellular carcinoma, leukemia and head-and-neck cancer models.
- Ferroptotic lipid-peroxidation signaling — CAP can suppress GPX4-dependent antioxidant protection through NRF2/GPX4-related pathways in susceptible tumors.
- Ca²⁺ signaling disruption — CAP-generated oxidants can mobilize intracellular Ca²⁺ from endoplasmic-reticulum and mitochondrial stores, contributing to senescence, mitochondrial stress and cell death in some models.
- Suppression of migration, invasion and EMT — frequently observed downstream of redox and survival-pathway perturbation.
- Therapy sensitization — preclinical evidence supports enhanced activity of cisplatin, temozolomide, mitomycin C, sorafenib, immune-checkpoint blockade and some physical therapies when combined with CAP.
Bioavailability / PK relevance: Conventional pharmacokinetic concepts such as plasma concentration, oral absorption, half-life and systemic bioavailability do not directly apply. CAP dose depends on device architecture, voltage and waveform, pulse characteristics, feed gas, gas flow, treatment distance, duration, treated area, tissue composition and delivered reactive-species flux. Short-lived plasma species largely act at or near the treatment interface, whereas longer-lived products such as H₂O₂, nitrite and nitrate can persist in plasma-activated liquids. Direct treatment therefore has an important tissue-penetration limitation, while plasma-activated liquids, endoscopic delivery and other carrier systems are being developed for deeper targets.
In-vitro vs systemic exposure relevance: CAP is not concentration-driven in the conventional drug sense, so comparison of an in-vitro molar concentration with an achievable human blood level is inappropriate. In-vitro results can nevertheless substantially overestimate clinical activity because monolayer cultures have little diffusion barrier and plasma dose depends strongly on medium volume, composition and treatment geometry. Translation therefore requires physical and chemical dosimetry rather than conversion to a systemic drug concentration.
Clinical evidence status: Strong preclinical evidence; early human evidence. A first-in-human Phase I study used intraoperative CAP following surgical resection in 20 patients with advanced or recurrent solid tumors and principally established feasibility and short-term safety. A separate Canadian first-in-human breast-cancer dose-escalation study, NCT06222788, is evaluating intraoperative NTP after lumpectomy and was recruiting in 2026. These studies do not establish CAP as standard cancer therapy. FDA authorization of an investigational trial or clearance of particular plasma devices for surgical uses should not be interpreted as general FDA approval of CAP for cancer treatment. Controlled efficacy trials and standardized CAP dosimetry remain major translational requirements.
Anticancer Mechanisms of Non-Thermal Plasma
| Rank |
Pathway / Axis |
Cancer Cells |
Normal Cells |
TSF |
Primary Effect |
Notes / Interpretation |
| 1 |
RONS and intracellular oxidative stress |
ROS ↑; RNS ↑; oxidative stress ↑ |
ROS ↑ or ↔ (dose-dependent) |
P/R |
Redox overload and tumor-cell injury |
Central CAP mechanism. H₂O₂, hydroxyl-related species, superoxide-related chemistry, NO-derived species, nitrite and nitrate vary substantially with device and medium. Cancer selectivity is frequent but not universal. |
| 2 |
Mitochondrial dysfunction and apoptosis |
Mitochondrial ROS ↑; membrane potential ↓; Bax ↑; Bcl-2 ↓; cytochrome c ↑; caspase-9/3 ↑; apoptosis ↑ |
Apoptosis ↔ or ↑ (dose-dependent) |
R/G |
Mitochondria-mediated cell death |
Well supported across multiple cancer models. Selectivity depends on plasma dose, baseline redox state and tissue type. |
| 3 |
Immunogenic cell death |
Calreticulin surface exposure ↑; HMGB1 release ↑; ATP release ↑; immunogenicity ↑ |
Context-dependent |
R/G |
Conversion of tumor-cell death into immune-stimulatory signals |
Provides rationale for combination with immune-checkpoint blockade and other immunotherapies. |
| 4 |
DNA oxidative damage and checkpoint activation |
8-oxo-dG ↑; DNA strand breaks ↑; γH2AX ↑; checkpoint signaling ↑ |
DNA damage ↑ (high dose); repair ↑ |
P/R/G |
Cell-cycle arrest, senescence or apoptosis |
Both direct oxidative DNA lesions and secondary damage from intracellular oxidative stress are reported. Persistent genotoxicity in surviving normal cells is not established and remains a safety consideration. |
| 5 |
Ferroptosis and lipid peroxidation |
Lipid ROS ↑; lipid peroxidation ↑; GPX4 ↓ (model-dependent); ferroptosis ↑ |
Generally lower sensitivity (model-dependent) |
R/G |
Iron-dependent oxidative membrane damage |
Increasingly well supported in breast, lung, endometrial, osteosarcoma and glioblastoma models. |
| 6 |
NRF2 and GPX4 antioxidant defense |
NRF2 ↓ or dysregulated; GPX4 ↓ (context-dependent) |
NRF2 ↑ or ↔ (context-dependent) |
R/G |
Weakening of antioxidant resistance and facilitation of ferroptosis |
NRF2 direction is not uniform across models. CAP can activate adaptive antioxidant signaling at sublethal doses but suppress NRF2/GPX4 protection in ferroptosis-sensitive tumor settings. |
| 7 |
PI3K AKT mTOR HIF-1α survival signaling |
AKT ↓; mTOR ↓; HIF-1α ↓ |
Variable |
R/G |
Reduced growth, hypoxic adaptation and survival signaling |
Strongly demonstrated in selected hepatocellular carcinoma, leukemia and head-and-neck models; not universal to every CAP configuration. |
| 8 |
Glycolytic metabolism |
Glycolysis ↓; glycolytic enzymes ↓ (model-dependent) |
Variable |
R/G |
Metabolic stress and reduced ATP-generating capacity |
Closely linked to suppression of AKT/mTOR/HIF-1α signaling in several cancer models. |
| 9 |
Ca²⁺ and mitochondrial ER signaling |
Cytosolic Ca²⁺ ↑; ER Ca²⁺ release ↑; mitochondrial Ca²⁺ dysregulation ↑ |
Ca²⁺ modulation ↑ or ↔ (context-dependent) |
P/R |
Senescence, mitochondrial dysfunction and death signaling |
Mechanistically meaningful but less universal than RONS. Melanoma studies support intracellular ER and mitochondrial stores as major contributors. |
| 10 |
Cell cycle and senescence |
G1 or G2/M arrest ↑; senescence ↑; proliferation ↓ |
Variable |
G |
Suppression of tumor expansion |
Checkpoint outcome depends on cell type, plasma dose and treatment format. |
| 11 |
Migration invasion and EMT |
Migration ↓; invasion ↓; EMT ↓ |
Usually limited at therapeutic exposure |
G |
Reduced metastatic phenotype |
Common downstream phenotype, although exact EMT signaling differs among cancer types. |
| 12 |
Chemosensitization |
Drug sensitivity ↑; resistance ↓ |
Context-dependent |
G |
Enhanced anticancer activity of combination therapy |
Preclinical synergy reported with cisplatin, temozolomide, mitomycin C, sorafenib and redox-targeting agents such as auranofin. |
| 13 |
Immunotherapy sensitization |
Antitumor immune susceptibility ↑ |
Immune activation may ↑ |
G |
Enhanced response to immune-checkpoint blockade |
Supported mainly by preclinical studies; likely related to immunogenic cell death and tumor-microenvironment modulation. |
| 14 |
Radiosensitization |
Radiation sensitivity ↑ (context-dependent) |
Uncertain |
G |
Potential enhancement of radiation-induced tumor injury |
Mechanistically plausible and supported in selected preclinical studies, but evidence is less mature than chemotherapy sensitization. |
| 15 |
Clinical Translation Constraint |
Response highly dependent on tumor geometry, redox phenotype and treatment access |
Toxicity ↑ if plasma dose is excessive |
P/R/G |
Limits reproducibility and clinical generalization |
No conventional PK metric exists. Major variables include plasma source, waveform, gas composition, flow, distance, duration, reactive-species dose, tissue penetration, direct versus plasma-activated-liquid delivery and tumor heterogeneity. Human evidence remains early-stage. |
P: 0–30 min R: 30 min–3 hr G: >3 hr
|