Science magazine has published the formal criticisms of the claim to have found extremophiles that substituted arsenic for phosphorus in their chemistry. It’s a thorough drubbing, and the most disappointing part is that Wolfe-Simon’s rebuttal simply insists that they were right, and doesn’t even acknowledge that many valid criticisms of the study were made. That’s not how you do it. Instead, she should answer the complaints by saying that they will do the experiments in a way that specifically addresses the perceived shortcomings of the study; she and her lab have their credibility invested in this research now, and the answer to the barrage is not to batten down the hatches and do nothing, but to do more experiments to show that the critics are wrong.

Nature also has a discussion of the issues, and this article bothered me in other ways. It rationalizes not doing anything to replicate or refute the work.

However, most labs are too busy with their own work to spend time replicating work that they feel is fundamentally flawed, and it’s not likely to be published in high-impact journals. So principal investigators are reluctant to spend their resources, and their students’ time, replicating the work.

“If you extended the results to show there is no detectable arsenic, where could you publish that?” said Simon Silver of the University of Illinois at Chicago, who critiqued the work in FEMS Microbiology Letters in January and on 24 May at the annual meeting of the American Society for Microbiology in New Orleans. “How could the young person who was asked to do that work ever get a job?” Silver said.

It’s true that this ought to be a relatively low priority for labs that are busy with good research, but it’s depressing to see that 1) whether something is publishable in high impact journals is such an important criterion for what we do, 2) skeptical science that replicates and refutes is considered a waste of effort, and 3) students are discouraged from carrying out such work, because there is some strange bias that will hurt their chances of employment.

I’m not disagreeing with those arguments, but I’m suggesting that they are symptoms of something rotten in the world of science. Testing claims ought to be what we do. If the journals are going to fill up with positive claims thanks to the file-drawer effect, and if nobody ever wants to evaluate those claims, and if negative results are unpublishable, the literature is going to decline in utility for lack of rigor and evaluation.

Of course, there is one group that has real incentive to get in there and get their hands dirty refining the results: the Wolfe-Simon lab. But her response implies she’s not going to make the effort (although I hope she really is doing something). And this attitude above suggests that, while the positive claim received a lot of media hoopla, any discovery of alternative explanations is going to get ignored. Methinks I see a ratchet at work.

I also notice that Rosie Redfield, brilliant genius that she is, has a relatively simple test of the claims. It’s not my field and I’m not equipped to do any of it, but I don’t see why anyone would find it a waste of effort to assign that project to a first-year grad student, as an exercise in techniques and skill, and as a way to get a quick (I know, low impact!) publication.

And I’m still bewildered that the scientific community would consider tests of a hypothesis a poor investment of its resources. This isn’t like creationism; Wolfe-Simon has a very specific claim that can be evaluated with standard laboratory techniques.

That study claiming that black women are “objectively” unattractive seems to be finally getting its author, Satoshi Kanazawa, in big trouble. That would be entirely wrong if it were based on disliking his conclusions, but if it were based on a demonstration of Kanazawa’s incompetence, then it would be earned. And that seems to be what is happening. Interestingly, many really good criticisms of Kanazawa are coming from Psychology Today’s blogs.

Daniel Hawes critiques his use of factor analysis and the problem of factor indeterminacy. This is a discussion of the failures of Kanazawa’s methodology, but also points out that he’s been peddling pseudoscience week after week.

An anticipated critique to what I’m saying here is that people will argue that I’m uncomfortable with his argument because it is politically incorrect. My above explanation has no reference to political correctness. The source of my frustration with Kanazawa’s writing is his pseudoscience. Given Kanazawa’s history of unabashedly blogging about research that he very well knows to be faulty at best and outright wrong at worst, my criteria for pseudoscience I discussed above are met. Yet, there is a natural reason that I (and others) have decided to respond to Kanazawa most recent article, and not as extensively to previous ones, that clearly follow the same disturbing pattern. Every other week there is a ridiculous Fundamentalist post claiming to explain “Why Night Owls are more intelligent”, or asking “Are all Women essentially Prostitutes?”, or posing the conjecture that “If Obama is Christian, Michael Jackson is White.” It seems hardly worth the effort to each time try to debunk the absurdity underlying these sensationalist arguments. However, when this unreasonable behavior spills into discussions of socially contentious issues such as race, I believe that pseudoscience left uncommented is dangerous. In particular it can quickly provide a basis for “scientific racism”, and so I believe that it is dutiful behavior for scientists and writers – especially when sharing the same media platform – to take a stance when these kind of discussions surface.

Scott Barry Kaufman and Jelte Wicherts have done something even more interesting: they downloaded the Add Health dataset that Kanazawa used and analyzed it independently. This is very revealing.

Kanazawa mentions several times that his data on attractiveness are scored “objectively”. The ratings of attractiveness made by the interviewers show extremely large differences in terms of how attractive they found the interviewee. For instance the ratings collected from Waves 1 and 2 are correlated at only r = .300 (a correlation ranges from -1.0 to +1.00), suggesting that a meager 9% of the differences in second wave ratings of the same individual can be predicted on the basis of ratings made a year before. The ratings taken at Waves 3 and 4 correlated between raters even lower, at only .136– even though the interviewees had reached adulthood by then and so are not expected to change in physical development as strongly as the teenagers. Although these ratings were not taken at the same time, if ratings of attractiveness have less than 2% common variance, one is hard pressed to side with Kanazawa’s assertion that attractiveness can be rated objectively.

The “waves” refer to the fact that the data is grouped by age into several categories, and he makes another interesting observation: if you look at only the adult wave, which is the only appropriate one if you want to talk about differences in sexual attractiveness, there are no differences by race.

Focusing just on Wave 4, it is obvious that among the women in the sample, there is no difference between the ethnicities in terms of ratings of physical attractiveness. Differences in the distributions for females when tested with a regular (and slightly liberal) test of independence is non-significant and hence can be attributed to chance (Pearson’s Chi-Square=15.6, DF=12, p =.210).

Now there’s the kind of statistical rigor I’ve come to expect and respect from my psychology colleagues.

So many people have sent me this sensationalistic article, “Scientists cure cancer, but no one takes notice“, that I guess I have to respond. I sure wish it were true, but you should be able to tell from how poorly it is written and the ridiculous inaccuracies (mitochondria are cells that fight cancers?) that you should be suspicious. The radical, exaggerated claims make the truth of the story highly unlikely.

Researchers at the University of Alberta, in Edmonton, Canada have cured cancer last week, yet there is a little ripple in the news or in TV. It is a simple technique using very basic drug. The method employs dichloroacetate, which is currently used to treat metabolic disorders. So, there is no concern of side effects or about their long term effects.

The simple summary is this: that claim is a lie. There have been no clinical trials of dichloroacetate (DCA) in cancer patients, so there is no basis for claiming they have a cure; some, but not all, cancers might respond in promising ways to the drug, while others are likely to be resistant (cancer is not one disease!); and there are potential neurotoxic side effects, especially when used in conjunction with other chemotherapies.

So we have one popular account that is badly written and makes exaggerated claims. There is also a university press release, the source for the sloppy popular account, that doesn’t contain the egregious stupidities but does tend to inflate basic research studies into an unwarranted clinical significance. And then, of course, there are the actual peer reviewed papers that describe the research and rationale, and also the reservations, on DCA. It’s like a game of telephone: you can actually trace the account from the sober science paper to the enthusiastic press release to the web account with its extravagant claims of a simple, cheap cure for cancer, and see how the story is gradually corrupted. It would be funny if the final result wasn’t going to dupe a lot of desperate people.

But there is a germ of truth to the story, in that DCA does have potential. Here’s how it works.

There are two major pathways that we use to extract energy from sugar. One is glycolysis, which extracts two ATP molecules from each molecule of sugar, and doesn’t require oxygen. Then there is glucose oxidation, which as you might guess from the name, does require oxygen, but which takes the byproducts of glycolysis and burns them completely to produce 36 ATP. So there’s the tradeoff: if your cells are oxygen-starved, or hypoxic, they can still get energy from sugar, but it’s relatively inefficient, but if they do have access to oxygen, they can extract much more. This is why you breathe, and why your heart beats, and why you have an elaborate circulatory system to deliver oxygenated blood to your tissues: without oxygen, you suffer a catastrophic hit to the efficiency of energy production.

Another feature of these two energy-producing pathways is that they are in different cellular compartments. Glycolysis takes place in the cytoplasm, while glucose oxidation occurs in the mitochondria. There is a gate-keeping enzyme, pyruvate dehydrogenase kinase (PDK), that regulates the flow of pyruvate, a product of the glycolysis pathway, into the mitochondria for oxidation. If PDK is active, it suppresses the transport of pyruvate into the mitochondria, and the cell is forced to rely on glycolysis, even if oxygen is available. If PDK is inactivated, pyruvate is shuttled into the mitochondria, even if oxygen is low.

This is where DCA comes in. DCA inhibits PDK, forcing cells to use the more efficient form of energy production. That sounds like a strange way to make a cancer cell uncomfortable, but the other factor here is that mitochondria are primary regulators of apoptosis, or cell suicide. They are loaded with sensors and enzymes that react to abnormalities in the cell (like being cancerous!) by activating a self-destruct mechanism. Shut down the mitochondra, you shut down the self-destruct mechanism that polices the cell. So the idea is a little indirect: by goosing the mitochondria, we also wake up the safety switch that, if all goes well, will cause the cell to spontaneously kill itself.

There are good reasons to think this might work. Many cancer cells arise in hypoxic environments; a poorly vascularized tumor, for instance, is going to be oxygen starved in the absence of blood flow, and the inhibition of mitochondria may be a factor in their survival. There is a well-known phenomenon called the Warburg effect, in which cancer cells will rely on glycolysis even when oxygen is available, suggesting that they have suppressed their mitochondria.

DCA also seems like a relatively safe drug. It’s been used for a long time in patients with metabolic disorders, or with metabolic side effects from other problems.

A large number of children and adults have been exposed to DCA over the past 40 years, including healthy volunteers and subjects with diverse disease states. Since its first description in 1969, DCA has been studied to alleviate the symptoms or the haemodynamic consequences of the lactic acidosis complicating severe malaria, sepsis, congestive heart failure, burns, cirrhosis, liver transplantation and congenital mitochondrial diseases. Single-arm and randomised trials of DCA used doses ranging from 12.5 to 100 mg kg^-1 day^-1 orally or intravenously). Although DCA was universally effective in lowering lactate levels, it did not alter the course of the primary disease (for example sepsis).

This is encouraging. It means there is a body of work already published on the effects of DCA, which should simplify the process of moving it into clinical trials. The authors, however, very clearly indicate that it won’t be a magic bullet affecting all cancers, but that some are likely candidates.

Dichloroacetate could be tested in a variety of cancer types. The realisation that (i) a diverse group of signalling pathways and oncogenes result in resistance to apoptosis and a glycolytic phenotype, (ii) the majority of carcinomas have hyperpolarised/ remodeled mitochondria, and (iii) most solid tumours have increased glucose uptake on PET imaging, suggest that DCA might be effective in a large number of diverse tumours. However, direct preclinical evidence of anticancer effects of DCA has been published only with non-small cell lung cancer, glioblastoma and breast, endometrial and prostate cancer. In addition, the lack of mitochondrial hyperpolarisation in certain types of cancer, including oat cell lung cancer, lymphomas, neuroblastomas and sarcomas, suggest that DCA might not be effective in such cases. Cancers with limited or no meaningful therapeutic options like recurrent glioblastoma or advanced lung cancer should be on top of the list of cancers to be studied.

Notice that the only work done so far is preclinical: that means it has been tested in mouse models, tissue culture, but hasn’t really been tried in cancer patients yet. The authors come right out and say that, express some possible reservations about its effectiveness, and suggest what needs to be done next.

No patient with cancer has received DCA within a clinical trial. It is unknown whether previously studied dose ranges will achieve cytotoxic intra-tumoral concentrations of DCA. In addition, the overall nutritional and metabolic profile of patients with advanced cancer differs from those in the published DCA studies. Furthermore, pre-exposure to neurotoxic chemotherapy may predispose to DCA neurotoxicity. Carefully performed phase I dose escalation and phase II trials with serial tissue biopsies are required to define the maximally tolerated, and biologically active dose. Clinical trials with DCA will need to carefully monitor neurotoxicity and establish clear dose-reduction strategies to manage toxicities. Furthermore, the pharmacokinetics in the cancer population will need to be defined.

Do not rush out and buy DCA and gurgle it down as a cancer preventative. We don’t know that it works — the safe concentrations for you may not be sufficient to kill any cancer cells, and the concentrations needed to kill cancer cells may be so high that it will do horrible, unpredicted, and dangerous things to you (some work with patients with congenital mitochondrial disorders also revealed some degree of peripheral neuropathy, for instance). This is why we have clinical trials: to work out safe and effective doses, look for dangerous interactions with other drugs — and if you have cancer, you’re already on a complicated cocktail of drugs — and detect unexpected side effects.

We should be urging further investigation of this promising drug with the beginning of clinical trials, but it’s far too early to be babbling about “cancer cures”. There have been lots of drugs that look great in the lab and have excellent rationales for why they should work, but the reality of cancer is that it is complicated and diverse and there are many more pitfalls between a drug that poisons cancer cells in a petri dish and a drug that actually works well in the more complex environment of a human being.

One other factor that inflames the conspiracy nuts over this drug is that DCA is simple, dirt-cheap, and completely unpatentable — there is no economic incentive for a pharmaceutical company to invest a gigantic bucket of money in clinical trials, because there is no hope for a return on the investment.

This is why an independent academic community with research funded for knowledge rather than profit is so important, and really emphasizes why we cannot afford to privatize all biomedical research. The authors propose a plan for progressing without the involvement of the pharmaceutical industry.

Funding for such trials would be a challenge for the academic community as DCA is a generic drug and early industry support might be limited. Fundraising from philanthropies might be possible to support early phase I – II or small phase III trials. However, if these trials suggest a favourable efficacy and toxicity, the public will be further motivated to directly fund these efforts and national cancer organisations like the NCI, might be inspired to directly contribute to the design and structure of larger trials. It is important to note that even if DCA does not prove to be the ‘dawn of a new era’, initiation and completion of clinical trials with a generic compound will be a task of tremendous symbolic and practical significance. At this point the ‘dogma’ that trials of systemic anticancer therapy cannot happen without industry support, suppresses the potential of many promising drugs that might not be financially attractive for pharmaceutical manufacturers. In that sense, the clinical evaluation of DCA, in addition to its scientific rationale, will be by itself another paradigm shift.

I can’t blame the industry for not following up on this: a clinical trial costs millions of dollars, and even if DCA pans out, there is no profit at all to be gained from it. For this research, we have to turn to public support (they have an interest in better cancer treatments!) and to scientists and doctors themselves, who of course have a great personal interest in seeing their patients get better.

Michelakis ED, Webster L, Mackey JR (2008) Dichloroacetate (DCA) as a potential metabolic-targeting therapy for cancer. Br J Cancer 99(7):989-94.