Cancer prevention

Cancer biology · Experimental evidence · 24 October 2026

The conditions that let cancer progress

What Ibrahim-Hashim’s 2012 bicarbonate experiment, together with later ammonia studies, suggests about the biology of prevention.

A potentially dangerous cell still has to survive in a tissue. It must obtain resources, withstand local conditions and expand despite the defences surrounding it. This makes cancer progression a question about the relationship between altered cells and their environment.

Experiments with extracellular buffering and ammonia clearance make that relationship tangible. Changing metabolic conditions has altered tumour development, immune control and treatment response in animal models. Taken together, these findings support a biological proposition: the conditions around abnormal cells can help determine whether those cells become progressive disease. [1] [4] [5] [6]

The central experiment · Ibrahim-Hashim et al., 2012

Changing the course of disease before a tumour appears

In Systemic Buffers Inhibit Carcinogenesis in TRAMP Mice, Ibrahim-Hashim and colleagues studied 37 genetically predisposed mice. Animals were randomized to control or four treatment groups, with 200 mM sodium bicarbonate in drinking water beginning at 4, 5, 7 or 10 weeks of age. [1]

Ibrahim-Hashim et al.: outcomes by treatment timing

Cohort	Mice	Reported outcome
Untreated controls	18	All developed ultrasound-detectable prostate tumours; mean detection age: 13 weeks.
Bicarbonate before 6 weeks	10	All reached 76 weeks without imaging-detectable prostate cancer.
Bicarbonate after 6 weeks	9	Cancer development did not differ significantly from controls.

Swipe the table to read the outcomes.

Microscopy adds an essential distinction: seven early-treated mice had hyperplasia without cancer; three had small, well-differentiated cancer foci. The result supports suppression of overt progression despite residual abnormalities. It does not establish elimination of every malignant cell. [1]

The small groups and engineered model limit generalisation. Even so, this experiment puts an important possibility on the table: prevention can include keeping an abnormal tissue from progressing to dangerous disease.

Why the environment can change the outcome

Acidity can participate in the behaviour it accompanies. Estrella and colleagues subsequently tracked tumour invasion and surrounding extracellular pH in living mice. Invasion was concentrated in the most acidic regions. Bicarbonate raised the pH around tumours and reduced local invasion in their experimental model. This intervention strengthened the case that acidity contributed to invasion. [2]

The broader interpretation is ecological. Conditions that damage neighbouring tissue, favour certain cellular adaptations or make expansion easier can influence which abnormal populations succeed. A lesion also changes its surroundings as it grows. Cause and consequence can therefore become linked: a metabolic disturbance produced by a tumour can subsequently help that tumour advance.

That distinction matters for prevention. A factor does not have to cause the first genetic alteration to become a useful point of intervention later. Preventing invasion, maintaining tissue organisation and restraining expansion are biologically meaningful outcomes in their own right.

Timing is part of the mechanism

A later TRAMP study by Astigiano and colleagues also found that treatment beginning early delayed progression. Treatment begun after tumours were present produced a trend that did not reach statistical significance. The study also reported hydronephrosis in some bicarbonate-treated animals, underlining that biological activity and long-term tolerability need separate evaluation. [3]

The timing results are compatible with an opportunity to influence disease while its supporting conditions are still developing. They do not identify one irreversible transition or prove why later treatment was less effective. Differences in lesion size, exposure, adaptation and tissue structure would need to be tested.

Ammonia adds immune control to the picture

Ammonia studies approach the same broad question through different chemistry. In colorectal cancer models, Bell and colleagues found that ammonia impaired T-cell function. Ornithine treatment reduced tumour ammonia and tumour growth. Crucially, the growth benefit disappeared in T-cell-deficient mice and after depletion of CD8 T cells. [4]

That dependence on immune cells is informative. It supports a mechanism in which improving metabolic conditions allows host defences to work more effectively. Another ammonia-lowering intervention, glycerol phenylbutyrate, also reduced tumour growth in the study. [4]

These were experiments on tumour growth and control. They do not establish that ammonia clearance prevents a first cancer in healthy people. Their contribution to the thesis is more specific: a metabolic disturbance can weaken the biological systems that restrain malignant cells.

The immune effects extend beyond T-cell exhaustion. Luo and colleagues found that ammonia impaired dendritic-cell phagocytosis and lymphocyte stimulation; removing ammonia reversed dysfunction caused by tumour-conditioned medium. Their work included cell experiments, mouse models and observations involving human blood and cirrhosis. [11]

Domagala and colleagues subsequently showed that ammonia reduced mature perforin and impaired NK- and T-cell killing. Their peer-reviewed 2025 paper supersedes the 2023 preprint. Ammonia increased pH within acidic organelles, and low-dose perforin loss was reversible after washout. The compartment matters: reducing extracellular tumour acidity and preserving acidic immune-cell granules are different biological objectives. [12]

Ammonia can also support the cells that establish tumours

Elaimy and colleagues investigated another route in hepatocellular carcinoma. They found that the transporter SLC4A11 supported ammonia uptake and nitrogen metabolism associated with cancer stem-cell properties. Ammonia exposure increased tumour-initiating capacity in transplantation experiments; ornithine treatment reduced tumour burden and stemness-associated measurements in mouse models. [5]

Here, “tumour initiation” means that existing cancer cells established tumours after implantation. It does not demonstrate transformation of healthy cells by ammonia. The study also linked higher ammonia to greater liver-cancer incidence in patients with cirrhosis, but that human observation cannot show that lowering ammonia would prevent cancer. [5]

The experimental findings nevertheless widen the biological argument. Metabolic conditions can influence both a malignant population’s capacity to establish itself and the host’s capacity to contain it. Those are distinct routes through which the same tissue may become more permissive to disease.

Ammonia supplies building material and growth signals

Spinelli and colleagues traced ammonia nitrogen into glutamate and other amino acids in breast cancer models, including tumours in mice. Glutamate dehydrogenase (GDH) mediated this recycling, and its depletion prevented ammonia-stimulated growth in the tested cells. Ammonia could therefore contribute directly to tumour biomass. [7]

Cheng and colleagues identified a separate signalling role: ammonia promoted SCAP–Insig dissociation and SREBP-1 activation, stimulating lipid synthesis. Disrupting the ammonia-binding site on SCAP suppressed tumour growth in their models. The original experimental report appeared in Nature Metabolism in 2022; Cheng, Kelsey and Guo later summarised it in Genes & Diseases. [9] [14]

Lu and colleagues found another connection in MDA-MB-231 breast cancer cells: ammonium chloride activated HIF-1 and increased aerobic glycolysis, with findings consistent with mitochondrial uncoupling and a role for mitochondrial reactive oxygen species. This provides a possible route from ammonia exposure to altered glucose metabolism, demonstrated in a particular cell model. [10]

How far does the nitrogen explanation go?

Meng and colleagues maintained adapted Hep3B cells for more than five months in glutamine-free medium supplemented with 0.8 mM ammonium. Growth was slower than in standard conditions. Other nutrients remained present: ammonia replaced a missing nitrogen supply rather than serving as the cells’ sole food. [8]

Biochemically, GDH can incorporate free ammonium into glutamate through reductive amination of α-ketoglutarate. Transaminases instead transfer amino groups already carried by other molecules. Glutamine synthetase incorporates ammonium into glutamine. Bypassing uptake of extracellular glutamine can therefore involve rebuilding glutamate and glutamine inside the cell; their biosynthetic functions remain necessary. [7] [8]

Meng and colleagues proposed that nitrogen anabolism—the production of nitrogen-containing molecules needed for cell growth—could help drive glutaminolysis and make the Warburg effect a downstream consequence of biosynthetic demand. Their glucose-related results included reduced expression of metabolic genes under nitrogen-source restriction, consistent with coordination between nitrogen supply and carbon metabolism. However, gene expression does not directly measure glycolytic flux, so the proposed causal link remains a hypothesis. [8]

Consistent with that boundary, Lie and colleagues found substantial differences between cell lines in their ability to use ammonia during glutamine deprivation. [13] A broader causal test would need to separate nitrogen availability, growth arrest, signalling and carbon flux rather than infer their order from changes occurring together.

Could hypoxia and ammonia reinforce each other?

Hypoxia and ammonia may form a reinforcing metabolic circuit. Cantwell and colleagues showed that hepatocyte HIF-2α suppresses HNF4α and urea-cycle enzymes, with elevated liver and circulating ammonia in their mouse model. Hypoxia also suppressed these genes in isolated hepatocytes in a HIF-2α-dependent manner. This provides a mechanism through which hypoxia signalling can impair nitrogen disposal. [15]

Ammonia can, in turn, activate HIF-1α in cancer cells. Alongside Lu’s glycolysis findings, Kitajima and colleagues showed that ammonia-induced HIF-1α promoted survival during metabolic stress while slowing proliferation. [10] [16] Where impaired disposal and ammonia-dependent signalling connect within the same tissue, ammonia could help maintain a hypoxia-associated metabolic state. This raises the possibility that impaired clearance becomes part of the machinery sustaining disease.

The full feedback loop remains a hypothesis: the strongest evidence for impaired hepatic disposal involves HIF-2α, whereas the cited cancer-cell responses involve HIF-1α. HIF activation also does not by itself establish continuing oxygen deprivation or extracellular acidity. Demonstrating that this circuit stabilises an acidic state would require measuring oxygen, nitrogen disposal and extracellular pH together, and showing that interrupting the proposed feedback weakens its persistence.

The response is context-dependent. Wang and colleagues found that some hypoxic cancer cells reduced ammonia release by diverting glutamine nitrogen into secreted dihydroorotate. [17] Whether ammonia accumulates therefore depends on the balance between its production, assimilation and removal.

The shared idea is metabolic permissiveness. Acidity and ammonia have different chemistry and cellular effects. Across these models, metabolic conditions can support biosynthesis, activate growth-related programmes and weaken immune control. Together, the findings suggest several ways a tissue can favour malignant expansion; they do not demonstrate one universal sequence operating in every cancer.

The connection to treatment response

Faes and colleagues showed another consequence of this dependence on conditions. In their colorectal cancer models, acidity reduced responsiveness to rapamycin. In HT29 xenografts, reported growth inhibition was 71.9% with rapamycin, 54.4% with bicarbonate and 98.3% with the combination. [6]

Those figures describe growth inhibition relative to controls, not cure rates. Groups contained five mice, and bicarbonate began before tumour-cell implantation. The design leaves open how much of the benefit depended on changing conditions before implantation. It supports further testing of the interaction between metabolic context and treatment response.

Across these studies, a common question emerges: which conditions allow a lesion to progress, evade control or withstand treatment—and which of those conditions remain changeable?

What this changes about the idea of prevention

The useful distinction is between preventing the first abnormal cell and preventing the consequences of its expansion. Research can address both. To test the second possibility, it would need to follow emerging lesions over time, measure their local metabolic conditions and ask whether an intervention changes progression to invasive disease.

For ammonia, a particularly important unanswered question is whether impaired clearance creates a permissive environment before cancer emerges, and in which organs or risk groups. Evidence from established tumours cannot settle that sequence. For buffering, the corresponding challenge is to identify a tolerable intervention that produces a relevant tissue change and improves a meaningful human outcome.

Neither approach has established reliable cancer prevention in humans. The biological conclusion is nonetheless substantial: genetic vulnerability unfolds within a tissue, and that tissue’s metabolic conditions and defences can influence the course of disease. The research opportunity is to preserve those restraining conditions early enough that potentially dangerous cells fail to become progressive cancer.

Research and related reading

1. Ibrahim-Hashim A, et al. (2012). Systemic Buffers Inhibit Carcinogenesis in TRAMP Mice. Journal of Urology 188:624–631. DOI: 10.1016/j.juro.2012.03.113.
2. Estrella V, et al. (2013). Acidity Generated by the Tumor Microenvironment Drives Local Invasion. Cancer Research 73:1524–1535. DOI: 10.1158/0008-5472.CAN-12-2796.
3. Astigiano S, et al. (2017). Systemic alkalinisation delays prostate cancer cell progression in TRAMP mice. Journal of Enzyme Inhibition and Medicinal Chemistry 32:363–368. DOI: 10.1080/14756366.2016.1252760.
4. Bell HN, et al. (2023). Microenvironmental Ammonia Enhances T Cell Exhaustion in Colorectal Cancer. Cell Metabolism. DOI: 10.1016/j.cmet.2022.11.013.
5. Elaimy AL, et al. (2024). SLC4A11 mediates ammonia import and promotes cancer stemness in hepatocellular carcinoma. JCI Insight. DOI: 10.1172/jci.insight.184826.
6. Faes S, et al. (2016). Acidic tumor microenvironment abrogates the efficacy of mTORC1 inhibitors. Molecular Cancer 15:78. DOI: 10.1186/s12943-016-0562-y.
7. Spinelli JB, et al. (2017). Metabolic recycling of ammonia via glutamate dehydrogenase supports breast cancer biomass. Science 358:941–946. DOI: 10.1126/science.aam9305.
8. Meng M, et al. (2010). Nitrogen anabolism underlies the importance of glutaminolysis in proliferating cells. Cell Cycle 9:3921–3932. DOI: 10.4161/cc.9.19.13139.
9. Cheng C, et al. (2022). Ammonia stimulates SCAP/Insig dissociation and SREBP-1 activation to promote lipogenesis and tumour growth. Nature Metabolism 4:575–588. DOI: 10.1038/s42255-022-00568-y.
10. Lu Y, et al. (2019). Ammonia mediates mitochondrial uncoupling and promotes glycolysis via HIF-1 activation in human breast cancer MDA-MB-231 cells. Biochemical and Biophysical Research Communications 519:153–159. DOI: 10.1016/j.bbrc.2019.08.152.
11. Luo C, et al. (2014). Ammonia Drives Dendritic Cells into Dysfunction. Journal of Immunology 193:1080–1089. DOI: 10.4049/jimmunol.1303218.
12. Domagala J, et al. (2025). Ammonia Suppresses the Antitumor Activity of Natural Killer Cells and T Cells by Decreasing Mature Perforin. Cancer Research 85:2448–2467. DOI: 10.1158/0008-5472.CAN-24-0749. Peer-reviewed version of the 2023 preprint.
13. Lie S, et al. (2019). The ability to utilise ammonia as nitrogen source is cell type specific and intricately linked to GDH, AMPK and mTORC1. Scientific Reports 9:1461. DOI: 10.1038/s41598-018-37509-3.
14. Cheng C, Kelsey S, Guo D. (online 2022; issue 2023). Glutamine-released ammonia acts as an unprecedented signaling molecule activating lipid production. Genes & Diseases 10:307–309. DOI: 10.1016/j.gendis.2022.07.017. Author commentary on reference 9.
15. Cantwell MT, et al. (2026). Hepatocyte HIF2α Downregulates the Urea Cycle Through Suppression of HNF4α. Gastro Hep Advances 5(9):101039. DOI: 10.1016/j.gastha.2026.101039.
16. Kitajima S, et al. (2017). Hypoxia-inducible factor-1α promotes cell survival during ammonia stress response in ovarian cancer stem-like cells. Oncotarget 8:114481–114494. DOI: 10.18632/oncotarget.23010.
17. Wang Y, et al. (2019). Coordinative metabolism of glutamine carbon and nitrogen in proliferating cancer cells under hypoxia. Nature Communications 10:201. DOI: 10.1038/s41467-018-08033-9.