Raffaghello · Proceedings of the National Academy of Sciences of the United States of America 2008 · In vitro and animal controlled experiment · n=?

Starvation-dependent differential stress resistance protects normal but not cancer cells against high-dose chemotherapy.

Level 5 - mechanism / opinion, no new human data

Preclinical in vitro and animal experimental study without human subjects.

PubMed 18378900 · doi:10.1073/pnas.0708100105 · record verified 2026-08-26

What was done

Researchers tested whether short-term starvation confers differential stress resistance against oxidative stress and chemotherapy across three models: S. cerevisiae (wild-type, proto-oncogene mutants, or expressing oncogenic Ras2(val19)); primary glial cells alongside six rat and human glioma and neuroblastoma cell lines exposed to hydrogen peroxide or cyclophosphamide under low-glucose or low-serum conditions; and mice bearing injected neuroblastoma tumors treated with high-dose etoposide following short-term fasting.

What was found

Short-term starved yeast or mutants lacking proto-oncogene homologs showed up to 1,000-fold greater protection against oxidative stress or chemotherapy drugs compared to Ras2(val19)-expressing cells. Low-glucose or low-serum media protected primary glial cells but failed to protect any of the six malignant cell lines against hydrogen peroxide or cyclophosphamide. In vivo, short-term starvation provided complete protection to mice against high-dose etoposide without protecting injected neuroblastoma cells. Exact animal sample sizes and quantitative survival rates were not reported in the abstract.

Why it matters

This study established the preclinical concept of starvation-dependent differential stress resistance, suggesting fasting could selectively shield healthy host tissues from cytotoxic chemotherapy without reducing tumor vulnerability.

Limits

The study is entirely preclinical, using yeast, cultured cell lines, and mice, so findings cannot directly establish clinical safety or efficacy in humans. Specific animal sample sizes, fasting durations, chemotherapy dosages, tumor burden metrics, and confidence intervals were omitted from the abstract.

Cited by