The PAC2 Report on the 2010 Pediatric Cancer Caucus highlighted a discussion by Dr. Melissa Hudson - Director, Cancer Survivorship Division, St. Jude Children’s Hospital and Dr. Kevin Oeffinger - Director, Adult Long-Term Follow-Up Program, Memorial Sloan-Kettering Cancer Center about survivorship and long-term effects of treatment.   They presented information from the Childhood Cancer Survivor Study conducted by 30 participating institutions that looked at 20,000 kids and siblings.   One amazing statistic they provided was that 73% of survivors had at least one chronic health effect and 42% had a very severe or fatal effect! 


But this post is the follow up we promised on one eye-catching chart they presented: excess mortality.  It sounded so horrible.  We wanted to give our understanding of it, and then look at the data in a different way.   We don't like what we see....


A 2008 article titled “Cause-Specific Late Mortality Among 5-Year Survivors of Childhood C..." published in the Journal of the National Cancer Institute was reviewed and key points are summarized below.  The article includes the eye-catching chart that we referred to, shown below.   The chart compares overall survival of kids that have survived 5-years since diagnosis to survival of other kids without cancer.  The difference in survival probability is excess mortality.  Excess mortality is the result of recurrence of the original cancer, secondary cancer or other health effect resulting from the toxicity of treatment; chemo or radiation.


The entire report is focused on defining the outcomes for 5-year survivors in terms of excess mortality.   The report's key findings include:

  • Of the 20,843 5-year survivors, a total of 2,821 deaths were observed, equivalent to about 14% of the survivors
  • Survivors had 8.4 times higher mortality following their 5-year survival after diagnosis when compared with the general age-, sex-, and year-matched US population.   (A similar report from Britain showed that survivors experienced 11 times the number of deaths expected from the general population)
  • Females had 13.2 times the expected mortality and males had 6.7 times the expected mortality
  • The overall estimated probability of a 5-year survivor living 30 years from diagnosis is 82%.  Probabilities were 94% at 10 years and 88% at 20 years
  • Highest excess mortality was observed in non-ALL, non-AML, medulloblastoma or PNET, other CNS malignancy and Ewing sarcoma
  • Excess mortality resulted from recurrent disease (58%), secondary cancers (19%), circulatory disease (7%) and respiratory disease (3%).


This is the chart.  The graph shows the probability of survival for male and female cancer kids (bottom two lines) compared to male and female non-cancer kids (top lines) over time.  As you can see, the probability of survival, living, for cancer kids is significantly less over time.  


So, there's another finding to chew on: The estimated probability of a 5-year survivor living 30 years from diagnosis is 82%.


This means that, in the case of a 19 year old, if he or she lives through a cancer diagnosis for 5-years, they are still faced with about a 20% chance of dying from cancer or cancer related-causes before age 49.  Does a 45 year old dying from recurring childhood cancer or complications from the cure sound 'cured'?


Or, in the case of a baby diagnosed at birth, if he or she lives through a cancer diagnosis for 5-years, they are still faced with about a 20% cancer of dying from cancer or cancer-related causes before age 30!   Does a young person dying at 25 from recurring childhood cancer or complications from the cure sound 'cured?


And what does this mean to our understanding of the commonly quoted 'cure rate' of 80%?   We should note that disease cure rates are all generally communicated based on 5-year survival.   But aren't we really concerned about whole life outcomes of kids here? 


Here's the math, looking at a one year time-frame in the US:

  • 13,500 children are diagnosed
  • 80% of the total are cured = 10,800 kids survive 5-years
  • 20% of the total die = 2,700 kids die
  • 18% of the 10,800 who are cured will die prematurely within 30 years of diagnosis from recurrent or secondary cancer or treatment related health-effects = another 1,944 who die.


2,700 + 1,1944 = 4,644 of the 13,500 kids diagnosed who either die within 5-years or will die within 30 years.  That's 34% of the total.   So, is the real 'cure rate' 66%?


1 in 3 of America's cancer kids dies within 30 years of diagnosis.

Does that change the way we should view and convey the statement "four in five children are cured of cancer"?




Disclaimer: our views and evaluation of the data, and if you can correct any errors, please leave a comment.  Thanks.

Views: 3052

Reply to This

Replies to This Discussion

Some further thoughts from Will's Dad:

Defining survival


Coalition Against Childhood Cancer

PAC2 is proud founding member of the Coalition Against Childhood Cancer!


  • Add Photos
  • View All

Latest Activity

DarrenChaker posted a blog post
Apr 30
DarrenChaker posted a discussion
Apr 30
DarrenChaker updated their profile
Apr 30
DarrenChaker posted a status
"I think joining organizations which support cancer research and help kids is important! Darren Chaker"
Apr 30



  • Add Videos
  • View All

© 2018   Created by AJs Dad.   Powered by

Badges  |  Report an Issue  |  Terms of Service