February 12, 2020 Par John Horgan 12 février 2020
John Horgan is a science journalist who manages Stevens Institute of Technology 's Science Writing Center. We are going to try to summarize the key points of his article here. Summary by Cécile Bour, MD
Cancer medicine generates enormous revenues, but marginal benefits for patients.
According to the author, there is a huge gap between the sad reality of cancer medicine in the United States and the optimistic claims made by the cancer industry and its media catalysts.
Cancer has spawned a huge industrial complex, including government agencies, pharmaceutical and biomedical companies, hospitals and clinics, universities, professional societies, non-profit foundations and the media.
Cancer-industry boosters claim that research, screening and treatment investments have led to "incredible progress" and millions of "cancer deaths averted," as claimed on the American Cancer Society's home page, a non-profit organization that receives money from biomedical companies.
Cancer experts and the media often describe new treatments with laudatory terms such as "medical breakthrough," "revolutionary," "miraculous," ("innovative," often heard in France, Editor’s note).
Cancer centers “frequently promote cancer therapy with emotional appeals that evoke hope and fear while rarely providing information about risks, benefits, costs, or insurance availability.” (as in the French Cancer League and the numerous cancer institutes, Editor’s note). These highlighted therapeutic achievements concern particular forms or rare cancers, which, according to the author, are exceptional in the midst of a "litany of failures".
The best way to measure progress against cancer is to look at mortality rates, the number of people who succumb to cancer per unit of population per year.
As the average life span of a population grows (because of advances against heart and respiratory disorders, infectious disease and so on), so does the cancer mortality rate. To calculate mortality trends over time, therefore, researchers adjust for the aging of the population.
With this adjustment—which, to be kept in mind, presents cancer medicine in a more favorable light — mortality rates have declined almost 30 percent since 1991.
This trend, according to cancer-industry boosters, shows that investments in research, tests and treatments have paid off. What boosters often fail to mention is that recent declines in cancer mortality follow at least 60 years of increases. The current age-adjusted mortality rate for all cancers in the U.S. is just under what it was in 1930!
Editor’s note: See the book "Dépistage du cancer du sein, la grande illusion" published by Thierry Souccar Editor for a more detailed explanation of this mechanism concerning breast cancer.
In reality, the rise and fall of cancer deaths track the rise and fall of smoking, with a lag of a couple of decades. Smoking raises the risk of many cancers but especially of lung cancer, which is by far the biggest killer, accounting for more deaths than colon, breast and prostate cancer combined.
It was concluded that without reductions in smoking "there would have been virtually no reduction in overall cancer mortality in either men or women since the early 1990s.”
New treatments yield small benefits, big costs
Clinical cancer trials “have the highest failure rate compared with other therapeutic areas”.
Pharmaceutical companies keep bringing new drugs to market.
But one study  found that 72 new anticancer drugs approved by the FDA  between 2004 and 2014 prolonged survival for an average of only 2.1 months… A 2017 study concluded that “most cancer drug approvals have not been shown to, or do not, improve clinically relevant end points,” including survival and quality of life.
The authors worried that “the FDA may be approving many costly, toxic drugs that do not improve overall survival”.
Screening leads to overdiagnosis and overtreatment
The cancer industry, aided by celebrities who claim that tests saved their lives, has convinced the public that screening for cancer is beneficial. The earlier we can detect cancerous cells, the more likely it is that treatment will succeed, John Horgan explains here one of the most significant findings of the past decade which is overdiagnosis. Many people have cancerous or pre-cancerous cells that, if left untreated, would never have compromised their health. Autopsies studies shown  that many people who die of unrelated causes harbor cancerous tissue.
Tests cannot reliably distinguish between harmful and harmless cancers. As a result, widespread testing has led to widespread overdiagnosis, the flagging of non-harmful cancerous cells. Overdiagnosis leads in turn to unnecessary chemotherapy, radiation and surgery.
Gilbert Welch was one of the first who brought the overdiagnosis to the light and called it « unfortunate side effect of our irrational exuberance for early detection.”
Editors’s note : In France, Bernard Junod , an epidemiologist, teacher and researcher at the Ecole des Hautes Etudes de Santé Publique de Rennes, was, together with Dr Bernard Duperray, a whistleblower on over-diagnosis in France concerning breast cancer screening. Mammograms and prostate specific antigen (PSA) tests have led to particularly high rates of over-diagnosis and over-treatment of breast and prostate cancer.
Adding together the harmful and lethal effects of screening, any benefits of screening "are outweighed by the deadly damage due to over-diagnosis and false positives" according to breast cancer specialist Michael Baum , co-founder of the screening programme in the UK, who is currently calling for the discontinuation of these programmes, which he believes could shorten more lives than they prolong.
For one man whose life is prolonged, many others will experience false positive results with subsequent additional tests, eventual prostate biopsy, over-diagnosis and over-treatment, and treatment complications such as incontinence and erectile dysfunction.
The prostate specific antigen discoverer, pathologist Richard Ablin, called the PSA test a "profit-motivated public health disaster."
All-cause versus specific mortality and « torturing the data »
Studies of tests for a specific cancer generally look at mortality attributed to that cancer. Mammograms are thus deemed effective if women who get mammograms die less often from breast cancer than women who do not get mammograms. This method can overstate the benefits of tests, because it might omit deaths resulting, directly or indirectly, from the diagnosis. After all, surgery, chemotherapy and radiation can have devastating iatrogenic effects, including heart disease, opportunistic infections, other forms of cancer and suicide.
Rather, reference should be made to studies that measure "all-cause" mortality, as they account for the deleterious effects of treatment. A 2015  meta-analysis by epidemiologist John Ioannidis (renowned for bringing the scientific replication crisis to light) and others found no reductions in all-cause mortality from tests for cancer of the breast, prostate, colon, lung, cervix, mouth or ovaries for asymptomatic patients.
In a recent editorial in the European Journal of Clinical Investigation, Ioannidis and four co-authors argue that cancer screening (especially mammograms and PSA tests) does more harm than good and should be abandoned. 
Often are pointed out improvements in survival rates, the length of time between diagnosis and death.
Survival rates for some cancers have indeed grown as a result of more widespread and higher-resolution testing, which detects cancer earlier.
But this doesn’t mean that people live longer as a result of early detection. They simply live longer with a diagnosis of cancer, with all its harmful emotional, economic and physiological consequences. (Explanation of survival rate here )
Using survival rates to promote tests is an example of what is have called “torturing the data» and for this article’s author, this a case of monstrous malpractice.
Corruption in the cancer industry
At USA, according to oncologist Vinay Prasad (@VPrasadMDMPH ; very active on twitter, Editor’s note) many cancer specialists accept payments from firms whose drugs they prescribe.
This practice "leads us to celebrate marginal drugs as if they were game-changers," Prasad argues. "It leads experts to ignore or downplay flaws and deficits in cancer clinical trials. ”
The desire of oncologists to produce monetizable findings might also compromise the quality of their research. A 2012 study of 53 “landmark” cancer studies found that only six could be reproduced.
Solution? Gentle cancer medecine?
The author pleads in favor of a medecine that adopt new therapies “when the benefit is clear and the evidence strong and unbiased.”
He pleads in favor of what some will call as conservative medicine that defines itself as a medicine that resists what is commonly and historically established, that knows how to say "stop" and arm itself with a lot of patience for those who want more medicine in a time where nobody is in favor of this attitude.
This meaning less testing, treatment, fear-mongering, military-style rhetoric and hype.
It would recognize the limits of medicine, and it would honor the Hippocratic oath: first, do no harm.
The patients must help these conservative physicians. We must recognize the limits of medicine and the healing capacities of our bodies, (dormant, non-evolving, spontaneously regressive cancers).
We must resist tests and treatments that have marginal benefits, at best.
We may never cure cancer, which stems from the collision of our complex biology with entropy, the tendency of all systems toward disorder.
But if we can curtail our fear and greed, our cancer care will surely improve.
According to Horgan, the fact to recognize the human body’s “inherent healing properties" and to acknowledge “how little effect the clinician has on outcomes” will make that physicians will thus protect themselves against their greatest foe—hubris, which is vanity.
 Extract from book of Bernard Duperray
Epidemiologists have the choice between two reference populations, the population of Europe or the world. The choice of the standard (“Europe” or “World”) induces wide variations in mortality.
“World” rates based on a younger population are generally lower than “Europe” rates reflecting an older population.
Depending on the standard chosen, the rate for the same population appears more or less high.
In France, the rates used most often by the InVs are the «Monde» rates (purple curve on the figure). The «Monde» standard tends to minimize the mortality rate for a country like France because it does not correspond to the structure of the French population. Moreover, according to Bernard Junod, age standardisation is subject to caution in describing a variation in mortality when the variations for the same age group are not uniform.
To assess the evolution of mortality in France in the fairest possible way, Bernard Junod chose to take as «standard» the age distribution of the population residing in France in 1992, that is to say in the middle of the period studied (1980 to 2005).
It thus obtained a standardized annual mortality rate by age in France per 100,000 women of 32.6 in 1980 and 32.9 in 2005, thus stable over this period.
In summary, here is what can be unquestionably retained from Figures 13 and 14:
• Between 1950 and today, the annual number of deaths and the crude rate of mortality from breast cancer have not decreased;
• Standardized rates “World” (purple curve fig 14) and “Europe” (blue curve fig 14) continuously increase in France until 1993 and then decrease while the crude rate stabilizes.
Let us now look again at figure 14. A surprising fact concerns us. The standardised “World” and “Europe” mortality rates of today, while falling, are still higher than those of the 1950s. Some would argue that comparing the current rates with those of 1950 is questionable because of the unreliability of the data at that time. It is possible, but then what about the fact that the standardized rates in 2006 were at the same level as those in 1970 according to the death registry which refers since 1968 (date of the creation of CépiDc, the Centre for Epidemiology on Medical Causes of Death)?