Cancer Mortality Continues to Decline Steadily

The trend for cancer, like the trend for longevity, is heading slowly in the right direction. Large investments in research produce incremental reductions in cancer mortality, but the shape of this relationship results from the present dominant approaches to cancer treatment, producing therapies that are each limited in their application to only one or a few narrow categories of cancer. Cancer is a spreading tree of variants, and all of the outer branches differ from one another in the details of their cellular biochemistry. Researchers tend to focus on attacking the particular distinctive biochemistry of one branch. This is inefficient and expensive, but it is going to change. The future of cancer treatment will hinge on approaches under development that are capable in principle of application to near all cancers, in particular methods of interfering in the lengthening of telomeres that all cancers rely upon. Once those treatments are a going concern, reduction in cancer mortality will no longer be an incremental trend.

Every year, the American Cancer Society estimates new cancer cases and deaths in the U.S. for the current year and compiles the most recent data on cancer incidence, mortality, and survival. Steady reductions in smoking combined with advances in cancer prevention, early detection, and treatment have resulted in a 23% drop in the cancer death rate since its peak in 1991. Overall cancer incidence is stable in women and declining by 3.1% per year in men (from 2009-2012), with one-half of the drop in men due to recent rapid declines in prostate cancer diagnoses as PSA testing decreases. Cancer mortality continues to decline; over the past decade of data, the rate dropped by 1.8% per year in men and 1.4% per year in women. The decline in cancer death rates over the past two decades is driven by continued decreases in death rates for the four major cancer sites: lung, breast, prostate, and colon/rectum.

Death rates for female breast cancer have declined 36% from peak rates in 1989, while deaths from prostate and colorectal cancers have each dropped about 50% from their peak, a result of improvements in early detection and treatment. Lung cancer death rates declined 38% between 1990 and 2012 among males and 13% between 2002 and 2012 among females due to reduced tobacco use. Even as cancer remains the second leading cause of death nationwide, steep drops in deaths from heart disease have made cancer the leading cause of death in 21 states. Heart disease remains the top cause of death overall in the United States. "We're gratified to see cancer death rates continuing to drop. But the fact that cancer is nonetheless becoming the top cause of death in many populations is a strong reminder that the fight is not over. Cancer is in fact a group of more than 100 diseases, some amenable to treatment; some stubbornly resistant. So while the average American's chances of dying from the disease are significantly lower now than they have been for previous generations, it continues to be all-too-often the reason for shortened lives, and too much pain and suffering."



Don't expect any significant advances in cancer treatments anytime soon. Cancer is different for each patient, and the genetics of cancer can be different within a single patient, too. Tumors in the same patient can have different genomes. We're nowhere near being able to treat cancer in a personalized way, and our ability to discover and research new drugs is stagnating. Cancer is going to be a serious and deadly disease for a long time to come.

Posted by: MissKaioshin at January 14th, 2016 6:27 PM

So what are you suggesting Florin, that no advances in knowledge have been made since 1975? Or that knowledge is being steadily accumulated and cancer death rates should have a 1-1 inverse relationship with knowledge about cancer?

I'm no expert, but it seems that medical technology for particular diseases advances in step changes rather than at a constant pace. You could argue that we are still miles away from any breakthrough, but the new immunotherapies do look interesting.

Posted by: Jim at January 14th, 2016 6:58 PM

I would think with AI soon coming on-line within a decade or so (Watson a pre-AI?), gene therapy (CRISPER), and the genome being decoded and now low cost, that we would have better results treating/curing cancer moving forward. Plus, as Jim suggested, Immunology.

Posted by: Robert Church at January 14th, 2016 8:41 PM

@Florin: I don't see the 185 number in either of the two pages you link to. has it at 166 in 2012 and agrees with the 199 for 1975. That'd make it a 16% drop. Still tiny, but a lot better than 7%.

Posted by: ale at January 14th, 2016 10:39 PM

Jim, I'm suggesting that the reduction in the overall cancer death rate listed in the PR piece isn't as impressive as it may appear at first glance. While I don't discount the possibility of dramatic breakthroughs in non-WILT cancer treatments, I'm not holding my breath either.

Ale, I got 185 using simple math from the 595,690 figure listed in the press release. Even if we go with 166, the 16.58% decease still isn't a heck of a lot. And what's worse is that the cancer death rate peaked in 1991. I'd expect cancer treatments to be much better in 1991 than in 1975.

Posted by: Florin Clapa at January 15th, 2016 1:42 AM

Florin: Probably the death rates of the post are age-adjusted death rates. Our medicine is better (and so we have less cancer for a given age) but we live longer (so we have a greater risk of developing cancer).

Posted by: Antonio at January 15th, 2016 2:15 AM

See for example

Mortality rates for all cancers combined have decreased overall for all of the broad age groups in the UK since the early 1970s, except for those aged 80+.[1-3] European age standardised (AS) mortality rates in 50-59 year olds have decreased from 313 deaths in every 100,000 in 1971-1973 to 179 in every 100,000 in 2010-2012 – a drop in rates of 43%.

The decrease in mortality rates in people younger than 50 is slightly greater than in the 50-59 age group (more than 50%); however, this age group accounts for less than 5% of the total cancer deaths. In people aged 60-69, the decline in mortality rates has not been so pronounced at around 31%. Although there has been an overall decrease in mortality rates of more than 9% in people aged 70-79, this includes an increase in females of 4% and a decrease in males of 26% (data not shown). In people aged 80 and over there has been an overall increase in mortality rates of 29% from 1,550 in every 100,000 people in 1971-1973 to 2,000 in 2010-2012; however, rates have fluctuated over the time period.

Posted by: Antonio at January 15th, 2016 2:21 AM

The whole point of the SENS foundation is to keep people in a state of youth where cancer rates are very low or non existent anyway. Deal with the largest risk factor for cancer, age (or biological age) to be more correct.

Posted by: Jim at January 15th, 2016 3:24 AM

Antonio, if we look at the age groups of 60 and above which have the vast majority of cancer deaths, the cancer mortality rate hasn't improved by much overall.

60-69: 31% decrease
70-79: 9% decrease
80+: 29% incease

Posted by: Florin Clapa at January 15th, 2016 2:19 PM

Mortality rates don't mean much if not taken together with incidence, which is not dependent only on aging but on lifestyle factors too. For example, the incidence of lung cancer in males peaked in the early eighties and is still increasing in females. Also, it could be misleading to extrapolate from large age groups since it is conceivable that the mean age of the 80+ group of 1975 is lower than the mean age of the 80+ group of 2012, which would thus result in a lower incidence for the first group. Finally, mortality changes for other diseases affect cancer death rates. Therefore, more people dying of cancer at the oldest ages is likely one of the consequences of the dramatic drop in cardiovascular mortality that took place over the last four decades. A better indicator of progress in cancer treatment would be comparing survival in 1975 with survival in 2012 for the same disease stage in order to control for artificially long survivals due to early diagnosis without a real improvement in treatment effectiveness.

Posted by: Barbara T. at January 15th, 2016 2:43 PM

Personally, I don't think much of potential "confounders," because if cancer treatment was so dramatically more effective today, it would overwhelm all of them.

Posted by: Florin Clapa at January 15th, 2016 3:13 PM

Breast cancer mortality has declined by "only" 40% between 1971 and 2011. However, incidence doubled. Taking this into account, progress in undeniable. Again, a comparison of survival at the same stage would be more appropriate (perhaps some of the early stage cancers diagnosed after the introduction of screening were indolent), but for this it is sufficient to take a look at the two year survival rate for metastatic melanoma: 88% in a 2015 immunotherapy trial versus 15% in the 2000s. Moreover, survival plateaus around year 3, hinting at the fact that most of these remissions are long term cures.

Posted by: Barbara T. at January 15th, 2016 5:37 PM

I'm not denying that there has been progress on some cancers in some age groups, but the overall cancer mortality rate, which is different from survival rates (i.e., how long people live after a cancer diagnosis), has not improved much. This is just squaring the curve; less people die of cancer at earlier ages but more die at older ages. And this is why I'm not that impressed.

Posted by: Florin Clapa at January 15th, 2016 8:23 PM

Florin Clapa said: "less people die of cancer at earlier ages but more die at older ages"

That is evidence of the slow progress of antiaging medicine instead of the slow progress of anticancer medicine.

Posted by: Antonio at January 16th, 2016 6:59 AM

Florin, my point is that you need to look at incidence, survival, and disease stage too if you want to extrapolate something meaningful from death rates since on their own they do not mean anything. In the 80+ group they include 98 year olds who die of cancer after they get it (a new primary rather than a recurrence) for the third time simply because they are so old. Cancer treatment has improved a lot, but of course eventually the odds will be against you if you keep falling sick due to the fact that: a) your very old body is fertile terrain for cancer; b) your very old body may not withstand the treatment's side effects; and c) they my choose to not treat you because you have a very old body.

Posted by: Barbara T. at January 16th, 2016 11:33 AM

Sorry, but I don't buy any of your arguments. There has been dramatic reductions in heart disease in ALL age groups. Yeah, even in the 85 and over group which has experienced a 50% reduction in mortality since 1970. And without any major fluctuations. The same kind of progress should have happened to caner treatment, but it didn't.

Posted by: Florin Clapa at January 16th, 2016 1:47 PM

Florin, by that logic you could also argue that Gene Therapy will never work, as only one gene therapy has been approved to date (Glybera).

Again you seem to be assuming a linear relationship between progress to cures and results in the clinic. I'm not sure this relationship actually exists. CAR T Cells had been worked on for three decades before their first demonstration of efficiency against Leukemia in 2011. By your logic of "only falling mortality shows progress" nothing was achieved between 1981 and 2010.

Posted by: Jim at January 16th, 2016 9:14 PM

Jim, there hasn't been pressure or time to develop gene therapy to anywhere near the same extent as cancer or heart disease research, so it's not a good comparison to make.

Posted by: Florin Clapa at January 16th, 2016 10:10 PM

The mortality rate for Alzheimer's disease has increased eight fold in the last forty years but this obviously doesn't mean that treatment in 1970s was eight times better than now. You can't make inferences on medical progress just by looking at crude death rates. There are a host of other factors that need to be considered, and failure to do so leads to a gross misinterpretation of the data.

Posted by: Barbara T. at January 17th, 2016 8:26 AM

The decline in mortality is largely due to one factor, a decline in smoking tobacco. Modern medical treatment has little to do with the reported decline in mortality.

Siegel and colleagues estimated that out of 345,962 cancer deaths in 2011 among U.S. adults 35 and older, 167,805 were associated with smoking. ( )

Fewer than 15% of U.S. adults currently smoke, according to new preliminary government statistics from the first half of this year. This is even lower than a report released just last week that found 16.8% of adults were smoking in 2014. The 14.9% rate is the lowest reported since this survey began in 1997, when 24.7% of U.S. adults were smokers. ( )

Those who quit smoking are likely to add 10 years to their life.
( )

Posted by: Bill Sardi at January 17th, 2016 10:16 AM

@Bill: the effect of reduced tobacco smoking takes decades to be felt, so while the incidence of lung cancer has been decreasing in males (but not in females), the full benefits of this behaviour change have not yet been seen. Also, a major contributor to cancer is obesity, which has been increasing at breakneck speed over the past three decades. This has resulted in higher incidence rates for a number of cancers - as opposed to lower as one would expect from less smoking - but not higher overall date rates, which would be the case if there hadn't been significant progress in treatment. Bottom line is, changes in cancer mortality rates cannot be attributed to a single cause and even complex mathematical modelling does not provide definite answers. Also, because they depend on a large number of factors (incidence at different ages, changes in the coding system, aging of the population, composition of the oldest age group, performance of competing diseases, fluctuations in the population's access to treatment etc) cancer death rates do not measure actual medical progress. In order to assess this we need to look at other indicators such as long term survival at different disease stages and eventually "cure" rates for metastatic cancer, which is what we are starting to see with immunotherapy.

Posted by: Barbara T. at January 17th, 2016 11:25 AM

I agree with Barbara, death rates alone don't mean anything.

Posted by: Antonio at January 17th, 2016 11:46 AM

@Florin Clapa: Heart disease has declined as a cause of death, with the common wisdom being that the results result from about a 50/50 split between lifestyle and medical technology. See, for example:

Incidence rates are another question; much of the mortality reduction is due to improved treatment where there is incidence.

Posted by: Reason at January 17th, 2016 3:34 PM

First, it's important to note that the decline in cancer death rates reported here by NIH and the American Cancer Society is, explicitly, adjusted for the age structure of the population. However, the role of changes in other risk factors in altering incidence, and of early diagnosis in facilitating treatment, also impact death rates without (usually) affecting the actual ability of medical treatment to effect cure once disease of a given stage is picked up.

Notably, as highlighted by others already, the smoking prevalence in the population has been cut roughly in half over this time period, which affects both cancer and heart disease mortality rates. The effects of screening are also important variables at play in determining mortality rates that don't affect the intrinsic efficacy of the therapies, but rather the "extrinsic" effectiveness of early treatment. Screening in some cases also inflates survival rates artificially, by detecting some indolent cancers that would not actually have become life-threatening within the course of current life expectancies: this is an important factor in inflating survival rates for breast cancer, and has historically been a particularly important one for prostate cancer.

What we are really looking for is the cure or "true" case-fatality rate. Because of the impossibility of ascertaining whether every last cancer cell has been purged from the body (and, perhaps, whether a permissive environment for cancer is left behind to set the stage for a "new" disease rising out of the old one), as well as cancer's evolvability, it operationally impossible to establish true "cure" rates, which is why we rely on the imperfect surrogate of 5-year survival rates.

For what they are worth, the latest figures show a rise in 5-year survival rates from cancer of all sites, from 49% in 1975-1977 to 68% in 2004-2010 (see page 18). Again, though, much of this is attributable to early diagnosis and sometimes inflation by treatment of non-life-threatening "cancers," particularly for cancers of the breast and especially the prostate. And 5-year survival after diagnosis is decidedly not cure of the disease. This is even less so if one wants to project out to a future in which mortality from other age-related causes is dramatically curtailed by repairing the cellular and molecular damage driving degenerative aging and its diseases.

This is real progress, but clearly indicates that we have a long way to go, and does not represent a trajectory in which we would achieve survival rates high and durable enough in the next few decades to obviate the need for a permanent prevention and cure for cancer.

Posted by: Michael at January 17th, 2016 5:09 PM

Reason, according to your 2012 post, the heart disease death rate has decreased more as a result of medical tech than lifestyle (since pills can't be considered a lifestyle choice).

Michael, do you think cancer medical treatment has improved significantly or is it too hard to tell? For me, it still seems that mortality rates are the best way to at least get a rough approximation of progress for medical treatments for cancer and heart disease.

Posted by: Florin Clapa at January 18th, 2016 1:00 AM

Gosh, Florin — I think I just answered that in exhaustive detail!

Posted by: Michael at January 19th, 2016 5:57 PM

Michael, no, it's not at all clear to me what your position is.

Posted by: Florin Clapa at January 19th, 2016 7:39 PM

I'm not very impressed by age-adjusted mortality data, since they adjust very different populations. They are transforming past populations (young in average) into current populations (old in average), so they arrive at cancer rates for aged populations, which are poor at any era anyway. It would be better to compare cancer mortality/survival at fixed ages, like these:

(The change is more near 50% than 18% for non-octogenaries.)

Sadly, statistical data for fixed ages is much more difficult to find on the internet than the age-adjusted ones, only available for some specific types of cancer, and I even couldn't find 5year-survival data.

Posted by: Antonio at November 30th, 2017 5:30 PM
Posted by: Florin Clapa at November 30th, 2017 8:01 PM

Thanks for the data!

So, looking at the 5-year survival data, there has been around a 50% decrease in the people that didn't last 5 years for the 50-64 and 65-74 years ranges from 1975 to 2009. Not bad. Could be better, but not bad. I also see that there isn't much difference between 5-year survival and 10-year survival, at least from 1988 (Relative Survival By Survival Time).

Mortality decrease is somewhat less than for breast cancer in the UK, mostly for the 50-64 years range. I would not say that it is "almost unchanged" in neither country. There is a clear peak around 1995 and then a decrease of around 40% in the UK and in the 65-74 years range in the US. I think that peak is probably due to underdiagnosis in previous years (if not, how can you explain a lower mortality with more primitive medicine and worse lifestyles?). So that means that the data doesn't show an "almost unchanged" evolution but an "at least 40% decrease", IMO.

The incidence also shows the peak. The decrease in recent years could cast doubts about the efficiency of modern treatments (the decrease could be due to less smoking, etc.), but looking again at the 5-year survival data, it's clear that the mortality change isn't mostly due to lifestyle changes.

Posted by: Antonio at December 1st, 2017 2:54 AM

Antonio, mortality is the only reliable measure. Survival just gets inflated by earlier and earlier detection, which helps sometimes, like chopping off a breast early, but mostly doesn't seem to help as much as expected given the overall curve.

Posted by: ale at December 1st, 2017 6:51 AM

Mortality is worse, because it depends on incidence. For example, more smoking means more mortality, even for the same panel of treatments. 5-year survival is not perfect, but I think it's better. As for the above breast cancer graph, it says: "Cancer sites include invasive cases only unless otherwise noted." There is a separate graph for non-invasive breast cancers (with 100% survival). So the overdiagnosis is atenuated.

Posted by: Antonio at December 1st, 2017 9:02 AM

The stats seem to show that yes, you'll live a little longer with your cancer, but it will still kill you. I don't see how incidence is very relevant; it's pretty much the same.

Posted by: Florin Clapa at December 1st, 2017 12:02 PM

Rereading the old posts, I see that we are running in circles. I already provided then the statistics I was searching for yesterday. And Barbara also made some points that I repeated today.

Posted by: Antonio at December 1st, 2017 12:28 PM

So, in summary: Yes, progress in oncology has been slow for the very aged. No, that doesn't mean that progress in oncology has been slow for all patients, it has been pretty good for middle aged and awesome for children. Yes, it means that for an aged population like our own, global mortality rates don't seem to have changed so much. Yes, for the old we will need WILT to prevent them from dying from cancer, or a similarly powerful therapy.

Posted by: Antonio at December 1st, 2017 12:37 PM

Progress has been slow for the vast majority of cancer patients (50+ age ranges). And it's not all old info either: the stats I recently mentioned suggest that this slow progress stretches all the way to the 1950s, not just to the 1970s. Here's another chart showing mortality rates starting from 1930. Same result.

There's no way to tell to what extent survival has improved from the 10-year survival chart you referred to. And I'm not sure how reliable or comprehensive those 10-year survival stats are anyway.

Posted by: Florin Clapa at December 1st, 2017 2:29 PM

Again... that is age-adjusted data. Why repeating the same thing once more?

Posted by: Antonio at December 1st, 2017 4:16 PM

Well, you seemed to have been OK with age-adjusted breast cancer mortality stats, but for some unspecified reason, you don't like age-adjusted all-sites cancer mortality stats. You seem to want stats for "fixed ages." Do you mean age ranges? I've already provided those. Non-age-adjusted raw numbers? Here you go.

Anyway, I have yet to see ANY stat that indicates significant progress has been made at reducing overall cancer mortality. If there were any, we'd know about it by now.

Posted by: Florin Clapa at December 1st, 2017 9:55 PM
Comment Submission

Post a comment; thoughtful, considered opinions are valued. New comments can be edited for a few minutes following submission. Comments incorporating ad hominem attacks, advertising, and other forms of inappropriate behavior are likely to be deleted.

Note that there is a comment feed for those who like to keep up with conversations.