Science · Comparative Oncology

Peto's Paradox: Why Elephants Almost Never Get Cancer

The evolutionary puzzle at the foundation of Nightbox's approach to cancer therapy.

The paradox in a sentence

If cancer starts from a single cell going rogue, then animals with more cells should get more cancer. They don't. That's it. That's the paradox.

Richard Peto, an Oxford epidemiologist, first noticed the discrepancy in the 1970s. He was comparing cancer rates across species and body sizes and the numbers didn't add up. A blue whale has roughly a thousand times more cells than a human. By naive probability, whales should be riddled with tumors before they hit adulthood. Instead, whales get cancer at roughly the same — or possibly lower — rates than we do.

The same pattern holds across mammals. Elephants carry about 100 times our cell count and live 60-70 years, plenty of time for mutations to accumulate. Yet the best epidemiological estimates put lifetime cancer incidence in elephants around 4.8%, compared to somewhere between 17% and 25% in humans. Something is actively protecting these animals.

How nature solved it

The short answer is redundancy. Elephants didn't evolve one magic gene — they evolved multiple overlapping defense layers, each of which compensates if another fails.

The best-studied layer is TP53. Humans have two copies of the TP53 tumor suppressor gene (one functional, one allele). Elephants have at least 20 copies, most of them retrogenes — processed pseudogenes that got reinserted into the genome over evolutionary time and then came back to life. Abegglen and colleagues at the Huntsman Cancer Institute published this finding in JAMA in 2015, showing that elephant cells undergo apoptosis at roughly double the rate of human cells when exposed to DNA-damaging agents.

But TP53 redundancy isn't the whole story. In 2018, Vazquez, Resigned, and Lynch at the University of Chicago published what I think is the more interesting finding: a resurrected pseudogene called LIF6. LIF6 (leukemia inhibitory factor 6) had been dead for tens of millions of years in the elephant lineage. At some point, a TP53-responsive promoter element inserted upstream of the defunct LIF6 coding sequence. The gene woke up. Now it functions as a p53-activated mitochondrial kill switch — when DNA damage is detected, p53 turns on LIF6, and LIF6 goes to the mitochondria and punches holes in the outer membrane. The cell dies. Cleanly. Before it can become cancerous.

The engineering question

If elephants can resurrect a dead gene and wire it into a cancer-defense circuit, the natural follow-up is: can we borrow that circuit?

That question is what led to NKG2D-LIF6 — the chimeric construct at the center of the Nightbox program. The idea is to take the elephant's LIF6 kill switch and wire it to a human-compatible tumor-detection module. We chose NKG2D as the sensor because it recognizes stress ligands (MICA and MICB) that are present on tumor cells and largely absent from healthy tissue. The result is a single gene therapy construct: human detection up front, elephant killing mechanism in the back, packaged in an AAV9 vector for delivery.

Whether it actually works in a living system is an open question. All of our data is computational. The wet-lab program starts in Q3 2026 with syngeneic murine tumor models in colorectal, melanoma, and pancreatic cancer. If the in vivo data holds up, we go to a pre-IND meeting at the FDA.

Why this matters now

Peto's paradox sat in the evolutionary biology literature for decades without many people trying to engineer from it. That's changing for two reasons. First, the structural biology tools caught up — AlphaFold3 lets you predict how a chimeric protein folds before you synthesize it, which makes construct design dramatically faster and cheaper. Second, AAV manufacturing costs dropped by roughly an order of magnitude since 2019, meaning a bootstrapped company can get research-grade vector without a $10M manufacturing budget.

That's the window Nightbox operates in. The biology has been published. The tools are available. The manufacturing is affordable. The question is just: does the chimera fold, does it bind, does it kill tumor cells, does it spare normal tissue? Five questions, each of which can fail. That's what the next 12 months are about.

Key references

• Peto R. (1977). Epidemiology, multistage models, and short-term mutagenicity tests. Origins of Human Cancer.
• Abegglen LM, et al. (2015). Potential mechanisms for cancer resistance in elephants and comparative cellular response to DNA damage in humans. JAMA 314(17):1850-1860.
• Sulak M, et al. (2016). TP53 copy number expansion is associated with the evolution of increased body size and an enhanced DNA damage response in elephants. eLife 5:e11994.
• Vazquez JM, Sulak M, Chigurupati S, Lynch VJ. (2018). A zombie LIF gene in elephants is upregulated by TP53 to induce apoptosis in response to DNA damage. Cell Reports 24(7):1765-1776.

Written by Artem Shakin, founder of Nightbox LLC. Published 2026-04-30. This page is part of our open science effort under CC BY 4.0.