(Ivanhoe Newswire) -- Cancer breakthroughs may be deceptive to the public. A study, published in the open-access journal PLoS Computational Biology on October 20, questions the reasoning used in many highly cited cancer publications supporting the relevance of animal and test tube experiments to human cancer.

Due to ethical guidelines, most experimental cancer research is performed on animals or cells living in test tubes.

"People are not guinea pigs and research on human subject is---fortunately---heavily constrained by ethics guidelines," said Vincent Detour, PhD and co-author of the study.

"As a result most biomedical research is conducted on cells living in test tubes or on lab animals. While this frees scientists from some ethical constraints, test tubes and animals do not necessarily mimic human biology."

Relating these laboratory experiments to human biology is a major hurdle; leading researchers at the Université Libre de Bruxelles to question the validity of several claims about cancer-causing genes

Detour goes on to explain that many factors differ in laboratory tests, compared to true progression of cancer in humans. For example, human progression may be effected by immune response, diet and physical exercise of the patient.

Additionally, mice don't frequently develop spontaneous cancers. Therefore, scientists must expose the animals to toxic treatments, inducing gene mutation and cancer growth, or manipulate cells and inject them into "nude mice", those with no immune system.

"Test tubes and lab animals are absolutely central to the progress of medicine, but there is a long way to bridge results from lab research and applications in the clinic," Detours said.

The connection from lab to human is often made using the prognostic value of molecular markers. In other words, if a certain biological process is found to inhibit cancer progression in the lab, researchers then look for the genes that are switched "on" when that biological process is taking place.

Once identified, humans with cancerous tumors are analyzed. If found that patients with the genes turned "on" fare worse than those with the genes turned "off", then the biological process is said to be prognostic (i.e. predictive of cancer progression) and clinical trials are scheduled.

The authors raise questions about this reasoning by highlighting that most random molecular markers are prognostic in human breast cancer, but are not relevant to the progression of the disease.

"Well, we reduced this argument at absurdum by demonstrating that in breast cancer nearly any gene, and even more so set of genes, is associated to disease progression," Detours said.

To confirm their concerns, the authors studied 47 cancer markers published in leading scientific journals. Their results confirmed the markers' prognostic values, but found that 60% of them were no more predictive of breast cancer outcome than completely random markers.

"So the problem with the 47 reviewed studies is not that that the genes they use are not predictive, it is that they completely fail to recognize that nearly any combination of genes is. Ask a chimp to select a few genes, these will almost certainly be associated with disease outcome," Detours stated.

He went on to explain some examples of this failure that were found during his studies; including a set of genes, selected by comparing the blood of Japanese diabetics after postprandial laughter (laughter after eating), that according to current procedure, successfully predicts the recurrence of breast cancer in women.

"This is obviously nonsense, no serious oncologist believes that you can cure cancer by telling a joke after lunch, yet it follows strictly the argument above. So, if the argument is inclusive for postprandial laugher, why should it be conclusive for cancer stem cells and other fancy molecular pathways that excite oncologists?" Detour asked.

Although the study was completed in 2007, Detour said he found considerable challenges while attempting to get the study published in the same top scientific journals that published the flawed studies.

Detour concludes, "It [the study] questions the biological interpretation of prognostic value. I see it this way: smoke is a very useful sign that there is a fire, although it does not cause fire."

He adds that it is important to remember that reports of "new hope against cancer" will often be deceptive because they are about test tube or animal research and it is extremely difficult to demonstrate that they apply in humans.

SOURCE: PLoS Computation Biology, published online October 20, 2011; e-mail interview with Dr. Vincent Detour held on October 18, 2011