Cabbage makes the news
Listed below are three headlines taken from UK media sources:
How the humble cabbage can cut cancers BBC news web-site 15/08/2018
Eating kale and broccoli can help prevent colon cancer, study claims The Independent 14/08/2018
Another reason to eat your greens! Compound in kale, cabbage and broccoli protects against bowel cancer The Daily Mail 14/08/2018
These headlines appeared in response to a paper published in the highly rated journal Immunity by scientists from London’s Francis Crick Institute and Imperial College. A short press release was issued on 14/08/2018 and a news article was posted on the Imperial College web-site written by a member staff from Imperial’s publicity department. The article on Imperial’s web-site clearly states that this is an “early stage mouse study” i.e. that it is at the bottom of the pyramid or hierarchy of evidence in medical and public health decision making. This is decidedly not the impression created by the BBC or Daily Mail headlines above. The Imperial web-site article states that the health benefits of vegetables are well established but that many of the mechanisms are unknown which I would accept as a balanced reflection of current understanding.
General back ground – health benefits of vegetables
There is overwhelming evidence that people who eat more fruit and vegetables have a reduced risk of death and increased life expectancy. Most of the benefit seems to be gained by eating five daily (80g) portions hence the 5-a-day campaign. The supporting evidence is largely from epidemiological studies which can never prove cause and effect nevertheless almost all scientists accept that high fruit and vegetable intake probably causes an increase in life expectancy. The evidence that high fruit and vegetable intake is associated with reduced incidence and death from cardiovascular diseases is also clear and consistent. Evidence of reduced cancer incidence with high fruit and vegetable consumption is less clear cut. The European Prospective Investigation into Cancer and Nutrition (EPIC) is one of the largest and most sophisticated cohort studies ever conducted. It has half a million subjects spread across several different European countries and was designed specifically to try to identify dietary factors associated with increased or decreased risk of cancers. An analysis of the relationship between fruit and vegetable consumption and overall cancer risk after about 9 years of follow-up found only a very small decrease in overall cancer risk associated with higher fruit and vegetable consumption. Because of the uncertainty involved in correcting for confounding variables, like tobacco and alcohol use, the authors of this report suggested that this association should be interpreted cautiously.
What is it about cabbage in particular?
Cabbage is one of a large group of so-called Cruciferous vegetables that includes cabbage, broccoli, sprouts, kale, watercress, radish, horseradish and wasabi. The mooted health benefits of broccoli has been a frequent focus of media and “food guru” attention and it is on most lists of superfoods. Cruciferous vegetables produce a number of secondary metabolites called glucosinolates that contain both a nitrogen and sulphur. When the plant is damaged (e.g. chewed) enzymes within the plant convert these glucosinolates to a range of breakdown products that seem to be involved in the plants defences against infection and predation. Some of these breakdown products are toxic to insects and are also responsible for the bitter taste of some cruciferous vegetables which may also discourage some predators.
Paradoxically some of these gluosinolate breakdown products have been suggested as having potential health benefits for people especially in cancer prevention and maybe even its treatment. Sulforaphane extracted from broccoli or broccoli powder itself is sold as a dietary supplement as is indole 3 carbinol (I3R) the substance used in the study that generated the headlines listed earlier. Most of the studies involving these compounds have been studies, like the one from the Imperial and Crick Institute group, that are at the bottom of the evidence pyramid: studies with laboratory animals, isolated tumour cells or cell-free enzyme systems or studies testing the effects on blood risk markers in short term experiments with human volunteers. Unless the effects are large, epidemiological methods become less useful the more specific the effect one is testing e.g. is there an association between cruciferous vegetable or even broccoli intake and bowel or total cancer risk. A 2012 fact sheet from the US National Cancer Institute, concludes that there is no consistent evidence from cohort studies to link high intakes of cruciferous vegetables with reduced risk of prostate, breast, colorectal or lung cancer.
News that a company was beginning phase 2 clinical trials with a stable synthetic sulforaphane (called Sulforadex) for treating breast and prostate cancer attracted UK media attention in late 2014 (Guardian 2014) . No results from these trials have so far been published. The company web-site accessed on 17/08/2018 suggested that 44 patients had been recruited onto a phase 2a clinical trial of this compound in advanced metastatic breast cancer. Whilst anecdotal evidence of possible benefit for individual patients is briefly mentioned, final data is not expected until the end of the year. It is unusual for a food plant to be the source for a therapeutic drug. Even though there are dozens of precedents for important drugs being derived from plants, these are rarely from plants that are regularly eaten as food. We have probably learned to avoid eating plants that contain high amounts of pharmacologically active substances or have a bitter taste. This makes food plants a low probability source of potential new drugs. Many foods have been transiently hailed as potential drug sources but that promise is rarely fulfilled e.g. after more than 120 clinical trials of curcumin (from turmeric) for various purposes no RCT has so far had a successful outcome (Nelson et al, 2017).
The study that caused the headlines?
The original Imperial/Crick Institute paper in Immunity is a complex, highly technical paper that is difficult for a non-immunologist to read. Neither cabbage nor broccoli is mentioned directly in the paper although on page 6 there is a statement that:
“The notion that a diet enriched with green vegetables reduces the risk of cancer formation is widely distributed, although evidence from several clinical trials remains inconsistent”
None of the three references that they use to support this statement is a clinical trial and they are essentially reviews of largely epidemiological data. This lapse does not alter their conclusion that there is no consistent evidence to support the notion of a preventative role for cruciferous vegetables in preventing cancer.
The study used a number of experimental models including:
- Specific strains of mice including some that were engineered to be highly susceptible to bowel cancer
- Mice with chemically induced bowel cancers produced by exposure to a mutagenic agent (cancer producing chemical) followed by two exposures to an irritant intended to speed up tumour formation
- What they called colon organoids i.e. “test tube” cultured colons made from clumps of cultured colon cells generated from stem cells.
A protein receptor on gut cells called the aryl hydrocarbon receptor AHR plays a key role in preventing chronic colon inflammation and ultimately cancer. Mice engineered to be completely lacking in AHR died of overwhelming infection when deliberately exposed to pathogenic bacteria. I3C (from cabbage and broccoli) is one of a number of substances referred to as AHR ligands that bind to and activate this receptor and thus reduce inflammatory responses in the gut. Giving I3C to these mice lacking AHR does not affect the outcome when they are exposed to the pathogen. However, they also used mice engineered to have an AHR deficit but not a complete absence of the functional protein. When these mice were exposed to the pathogen they responded as those with no AHR but when fed an I3C enriched diet they were protected i.e. activation of AHR by dietary ligands like I3C protects the gut from a severe inflammatory response and overwhelming infection.
Mice with an incomplete AHR deficit rapidly developed bowel cancer when exposed to the chemical cancer-inducing regimen noted earlier. Wild type mice did not develop cancers with the doses of carcinogen used on those with the AHR deficit. Addition of I3C did prevent this chemical cancer induction in some mice with specific AHR deficits and even partially arrested developing cancer in a sample of 9 AHR deficit mice; they developed significantly fewer tumours which were more benign.
Using the lab grown colons they were able to show that without AHR the intestinal stem cells failed to develop into functional epithelial cells that were able to absorb nutrients and secrete protective mucous. Rather these cells grew uncontrollably which is the hallmark of cancer cells. Wild type mice with no genetic defect developed tumours within 10 weeks after exposure to the chemical cancer-inducing regime if they were fed a purified diet that the authors assumed was deficient in AHR-stimulating compounds but not if they were fed standard lab chow.
In these mice and in the lab grown colons, the AHR receptor plays an essential role in the development of functioning gut cells and in regulating their response to insults like pathogens and chemical carcinogens. Adding substances like I3C to diets and in vitro systems activates AHR and can reduce the harmful effects of these insults and reduce the intestinal damage and cancer development.
Some personal reflections
Very few of these fundamental studies, even those from high prestige institutions and published in top journals, end up being translated into new treatments for people. Many fall at the first hurdle and cannot be repeated in an industrial pharmaceutical laboratory or else they fail to give similar results when tested with other strains or species or ultimately people. It is very difficult to decide how relevant studies using a chemical derived from “cabbage” in specific types of mice are in when trying to decide how human bowel cancer risk is likely to be affected by normal variation in green vegetable consumption. How the experimental doses translate into daily portions of green vegetables is just one of many difficult, probably unanswerable, questions.
Evidence that glucosinolates in cruciferous vegetables like cabbage and broccoli have a specific role in preventing human bowel cancer is weak. This paper does not affect that conclusion nor was it ever intended to. This is not the message conveyed by the media headlines or some of the articles announcing its publication. The claims about potential holistic benefit for people in the formal scientific paper are muted and even the headline of the press release makes clear that these are findings in mice:
“Chemicals found in vegetables prevent colon cancer in mice”.
Some of the comments made in media interviews with the authors are less cautious. One author suggests that she has changed her diet and she gives advice on how to cook broccoli, presumably in order to maximise the content of these potentially protective chemicals. She implies that the study effectively shows how cabbage or broccoli prevent bowel cancer even though the evidence that they do so is weak.
In a recent post I noted that Scottish scientists were feeding farmed salmon on genetically engineered feed to increase the content of omega-3 fatty acids even though there is little evidence that omega-3 fatty acids are beneficial. Are plant breeders at this moment working to maximise the glucosinolate content of cruciferous vegetables or can we expect promotion of particular types or varieties of cruciferous vegetables because of their high glucosinolate content (and probably bitter taste)?