Cancer cells not only survive in oxygen but depend on it. Even the most oxygen-starved tumors contain regions where cells are bathed in oxygen and use it to fuel their growth. The real story is more nuanced than a simple yes or no: cancer cells are remarkably flexible in how much oxygen they need, thriving across a wide range of concentrations that would stress or kill many normal cells. That metabolic flexibility is central to what makes cancer so difficult to treat.
How Much Oxygen Tumors Actually See
A common misconception is that tumors are uniformly oxygen-deprived. They are not. A solid tumor is a patchwork of oxygen-rich and oxygen-poor zones, and the cells in each zone behave differently. Normal, healthy tissues sit at oxygen levels averaging around 5%, ranging from roughly 3% to about 7%. Tumors, by contrast, typically run lower, with median oxygen levels falling between about 0.3% and 4.2%, and most tumors sitting below 2%. Some cancers, particularly prostate and pancreatic tumors, are profoundly oxygen-poor.
1PubMed Central. Defining normoxia, physoxia and hypoxia in tumours-implications for treatment responseBut “lower than normal” is not “zero.” The cells closest to blood vessels within a tumor may see oxygen levels comparable to healthy tissue. Cells farther from a vessel see progressively less. And cells at the very center of a large, poorly vascularized tumor mass may be nearly anoxic. This gradient is not static; it shifts as blood vessels open and close, sometimes cycling on a timescale of minutes. The important takeaway is that many cancer cells live at perfectly serviceable oxygen levels, while their neighbors deeper in the tumor do not.
The Warburg Effect and Why Oxygen Levels Are Not the Whole Story
In the 1920s, Otto Warburg observed something strange: cancer cells consumed glucose and produced lactic acid at high rates even when plenty of oxygen was available. This “aerobic glycolysis” seemed wasteful, since burning glucose with oxygen yields far more energy than fermenting it without. Yet cancer cells do it anyway. The phenomenon, now called the Warburg effect, led to a widespread but oversimplified belief that cancer cells somehow reject oxygen or cannot use it.
That belief is wrong. Cancer cells engage in aerobic glycolysis not because they cannot use oxygen but because doing so offers other advantages. Rapid glucose breakdown generates the building blocks cells need to divide quickly, and it acidifies the surrounding tissue in ways that can suppress immune responses and promote invasion. The molecular machinery behind the Warburg effect typically runs through a protein called HIF-1α, a transcription factor that normally ramps up when oxygen is scarce. In cancer, oncogenes can switch HIF-1α on even when oxygen is plentiful.
2PubMed. The Mechanisms of Regulation of Aerobic Glycolysis (Warburg Effect) by Oncoproteins in Carcinogenesis Research on pancreatic cancer cells, for example, has shown that signaling through pathways like PI3K/Akt and Ras/MEK/ERK can drive HIF-1α accumulation under completely normal oxygen conditions, boosting the expression of genes that promote invasion.3PubMed. Stem cell factor/c-kit signaling enhances invasion of pancreatic cancer cells via HIF-1α under normoxic condition
So cancer cells in oxygen-rich zones are not sitting idle. They are actively metabolizing, dividing, and activating invasion programs, often using the same molecular switches that hypoxic cells use. Oxygen does not shut cancer down; cancer has learned to co-opt the body’s oxygen-sensing machinery regardless of how much oxygen is actually present.
Cancer Cells Still Need Their Mitochondria
If the Warburg effect were the whole picture, you might expect cancer cells to have ditched their mitochondria entirely. They have not. Even in cells that rely heavily on glycolysis, somewhere between 40% and 75% of ATP comes from that less efficient glucose-to-lactate pathway, while the rest is still produced in mitochondria through oxygen-dependent processes.4Biochimica et Biophysica Acta (BBA) – Bioenergetics. Oxidative phosphorylation in cancer cells That means cancer cells are not running entirely on fermentation. They are running a hybrid engine, and the oxygen-using part matters more than older models of cancer metabolism assumed.
This matters especially for two dangerous subsets of cancer cells: cancer stem cells and treatment-resistant cells. Both tend to depend heavily on mitochondrial respiration, the oxygen-consuming pathway, rather than glycolysis. Studies have consistently found that increased dependence on this oxygen-driven metabolism is a hallmark of cancer stem cells and of cells that survive chemotherapy or targeted therapies.5PubMed Central. Why All the Fuss about Oxidative Phosphorylation (OXPHOS)? These resistant cells promote mitochondrial activity through various signaling pathways and can be selectively killed by drugs that block mitochondrial respiration.6PubMed Central. The Effect of Oxidative Phosphorylation on Cancer Drug Resistance
The implication is counterintuitive: the cancer cells you most want to eliminate, the ones that resist treatment and can regrow a tumor, are often the most oxygen-dependent. They are not the glycolytic, Warburg-effect poster children. They are the ones quietly burning fuel in their mitochondria.
The Metabolic Symbiosis Inside a Tumor
Tumors are not just collections of identical cells. The oxygenated cells near blood vessels and the hypoxic cells farther away actually cooperate metabolically, forming a kind of internal economy. Hypoxic cells, running on glycolysis, churn out lactic acid as a waste product. That lactate diffuses outward through the tissue. Oxygenated tumor cells, rather than competing with their hypoxic neighbors for glucose, pick up that lactate and burn it in their mitochondria as fuel. This frees up glucose for the oxygen-starved cells deeper in the tumor.
7PubMed Central. Targeting lactate-fueled respiration selectively kills hypoxic tumor cells in miceThis symbiosis means the oxygenated cancer cells are not just surviving in oxygen, they are performing a critical support role for the entire tumor. Their oxygen-fueled metabolism recycles the waste products of their hypoxic neighbors, making the whole system more efficient than it would be if every cell were competing for the same nutrients. Disrupting that cooperation, by blocking lactate uptake in oxygenated cells for instance, can starve the hypoxic cells of glucose and kill them. This is one reason researchers have explored targeting the metabolic handoff between different tumor zones rather than just trying to kill all cancer cells with the same drug.
Why Oxygen Can Also Be Dangerous to Cancer Cells
There is a flip side to cancer’s oxygen use. Oxygen-driven metabolism generates reactive oxygen species (ROS), unstable molecules that can damage DNA, proteins, and the fatty membranes surrounding cells. Normal cells have built-in antioxidant defenses to handle routine ROS production. Cancer cells, which often have ramped-up metabolism and genetic instability, tend to produce more ROS than normal and therefore lean harder on antioxidant systems to survive.
This dependence on antioxidants has emerged as a vulnerability. When researchers strip away antioxidant defenses in cancer cells, the resulting buildup of ROS can trigger a form of cell death called ferroptosis, which is driven by iron-dependent damage to membrane fats. High levels of intracellular ROS from oxidative stress can make cells more susceptible to ferroptosis or accelerate its onset.8PubMed Central. Role of oxidative stress-induced ferroptosis in cancer therapy The reliance of cancer cells on antioxidant activity in certain contexts represents a potential vulnerability that could be exploited therapeutically.9PubMed Central. Antioxidant Defenses: A Context-Specific Vulnerability of Cancer Cells
Some experimental compounds work by pushing cancer cells over this oxidative edge. Timosaponin AIII, a plant-derived compound tested against colorectal cancer cells, was shown to trigger the breakdown of lipid droplets, releasing fatty acids that underwent oxidation. This cascade led to mitochondrial shrinkage and a drop in key ferroptosis-protective proteins, ultimately killing the cells.10PubMed. Timosaponin AIII induces lipid peroxidation and ferroptosis by enhancing Rab7-mediated lipophagy in colorectal cancer cells The strategy essentially turns cancer’s own oxygen-using metabolism against it by removing the safety net that keeps oxidative damage in check.
What Happens at Zero Oxygen
If cancer cells can handle low oxygen and use normal oxygen, what about no oxygen at all? This is where their flexibility runs out. Research has shown that while tumor cells can grow in hypoxia, they cannot grow in true anoxia. Molecular oxygen is not just fuel for the mitochondria; it is a required ingredient for several cellular processes that have nothing to do with energy production. Cells need oxygen to make cholesterol and other sterols, to fold proteins correctly, and to run a variety of enzymes including hydroxylases and histone demethylases that regulate gene activity.11PubMed Central. Oxygen Consumption Can Regulate the Growth of Tumors, a New Perspective on the Warburg Effect
These oxygen-dependent functions work fine even at quite low oxygen levels, which is one reason cancer cells can proliferate in the hypoxic interior of a tumor. But in the complete absence of oxygen, none of them can proceed. This anoxic death has been observed both in lab dishes and in living tumors. It draws a hard biological line: cancer cells are adaptable, but they are not anaerobic organisms. They need at least trace amounts of oxygen to sustain the non-energy biosynthetic work that cell division requires.
How Oxygen Levels Shape Treatment Outcomes
Oxygen’s role in cancer therapy is enormous and often underappreciated by patients. Radiation therapy, one of the oldest and most widely used cancer treatments, works in large part by generating DNA-damaging free radicals from oxygen. The amount of DNA damage radiation inflicts depends primarily on how much oxygen is in the immediate environment around the DNA, not on the type of cell being irradiated.12PubMed. Radiation-induced DNA damage in tumors and normal tissues: IV. Influence of proliferation status and cell type on the formation of oxygen-dependent DNA damage in cultured cells Hypoxic cancer cells, therefore, are inherently more resistant to radiation. They receive the same dose but suffer less damage because there is less oxygen around to amplify the effect.
This oxygen-dependent gap in radiation effectiveness has driven decades of research into ways to oxygenate tumors before treatment, or to design drugs that exploit low oxygen rather than fighting it. Hypoxia-activated prodrugs, for example, are compounds designed to remain inactive in oxygenated tissue but convert into toxic drugs inside hypoxic tumor zones. The number of preclinical and clinical trials exploring these agents has been steadily growing.13PubMed Central. Molecular Pathways: Hypoxia-Activated Prodrugs in Cancer Therapy
On the other side of the coin, anti-angiogenic therapies, drugs that try to starve tumors by cutting off their blood supply, can backfire. Reducing blood flow to a tumor lowers its oxygen levels, and the resulting hypoxia can trigger escape mechanisms. In glioblastoma, for instance, anti-angiogenic treatment-induced hypoxia has been linked to increased tumor invasion, shifts toward more glycolytic metabolism, activation of survival pathways, and enhanced self-renewal of cancer stem cells.14PubMed Central. Hypoxia-Mediated Mechanisms Associated with Antiangiogenic Treatment Resistance in Glioblastomas The tumor adapts to oxygen deprivation and becomes harder to treat.
Tumors Build Their Own Blood Supply
Cancer cells do not passively accept whatever oxygen drifts their way. Tumors actively recruit new blood vessels through a process called angiogenesis. As hypoxic zones develop inside a growing tumor, cells there ramp up production of vascular endothelial growth factor (VEGF), a signaling protein that stimulates nearby blood vessel cells to sprout new capillaries toward the tumor.15PubMed Central. Tumor angiogenesis: causes, consequences, challenges and opportunities The new vessels are often leaky and disorganized, which is partly why oxygen distribution within tumors is so uneven, but they deliver enough blood to sustain growth at the tumor’s expanding edge.
This vascular recruitment means that a tumor’s oxygen supply is not fixed. A small tumor with a limited blood supply might be uniformly hypoxic, but as it grows and triggers angiogenesis, some regions become well-oxygenated while others fall further behind. The result is the mosaic of oxygen-rich and oxygen-poor zones described earlier. Researchers have even found that pharmacologically increasing a tumor’s oxygen consumption, by blocking HIF-1 or its downstream targets so that cells burn more oxygen in their mitochondria, can actually make the tumor more hypoxic overall by depleting the available supply faster.16Proceedings of the National Academy of Sciences. Metabolic targeting of hypoxia and HIF1 in solid tumors can enhance cytotoxic chemotherapy The interplay between supply and demand is constant, and both sides are moving targets.
Cancer Stem Cells and Their Metabolic Flexibility
Cancer stem cells, the small subpopulation thought to be responsible for tumor regrowth after treatment, add another layer of complexity to the oxygen question. As noted earlier, these cells tend to rely on mitochondrial respiration more than the bulk of the tumor. But they are also metabolically flexible, able to switch between glycolysis and oxidative metabolism depending on their environment.
Imaging studies in glioblastoma have highlighted these differences. When tumors grown from a standard glioblastoma cell line were irradiated, researchers observed a measurable drop in glycolytic activity, consistent with radiation damaging the cells’ metabolic machinery. But tumors grown from cancer stem-like cells showed no significant change in glycolytic metabolism after the same radiation dose, suggesting these cells either maintained their metabolic program or adapted quickly.17Cancer Research. Detection of metabolic change in glioblastoma after radiotherapy using hyperpolarized 13C-MRI This resilience in the face of treatment is part of what makes cancer stem cells such a challenge, and it underscores that their relationship with oxygen is not a simple dependence or independence but a dynamic adaptation.
Why Lab Conditions Mislead Researchers
One reason the role of oxygen in cancer has been so difficult to pin down is that most laboratory experiments on cancer cells have been conducted at atmospheric oxygen levels, around 20%. That is roughly four times higher than what normal tissues experience and far above what most tumors see. Growing cancer cells at atmospheric oxygen amounts to studying their behavior in an environment that does not exist anywhere in the human body.
18PubMed Central. Supraphysiological Oxygen Levels in Mammalian Cell Culture: Current State and Future PerspectivesThis mismatch has real consequences. Cells grown at 20% oxygen show different drug sensitivities, different hormone responses, and different metabolic profiles than the same cells grown at physiologically realistic oxygen levels. Research has demonstrated that hyperoxic culture conditions can introduce artifacts that distort how cells respond to treatments, leading to findings that do not hold up when tested in living organisms. There has been growing recognition in the field that maintaining realistic oxygen levels in the lab is necessary to produce results that actually translate to patients. But changing decades of standard practice is slow, and many published studies on cancer cell behavior were conducted under conditions that may have skewed the results in ways we are still sorting out.
This point has practical implications for how you interpret cancer research headlines. A study announcing that a particular compound “kills cancer cells in the lab” may have tested those cells at oxygen levels no tumor cell would ever encounter inside a patient. The cancer cells survived or died under conditions that bear little resemblance to the tumor microenvironment. That does not mean the finding is useless, but it does mean the road from lab dish to clinic is longer than it might appear.