Clinical Aspects of Cancer – Wim Ceelen
1. Cancer: epidemiology and risk factors
(Point) prevalence= proportion of existing cases (old and new) in a population at
a single point in time (cross-sectional)
Incidence= amount of new cases during a specified time interval (longitudinal)
- Risk= N cases/total population/t
- Odds= N cases/diseases free/persons over the study period => (risk/1-
risk)/ t
- Incidence rate= N cases/total person-years/t
2. Staging – TNM system
2.1 Measures of cancer survival and mortality
Overall survival: all cause, observed or crude survival => includes other causes
of death so it is an overestimation
Net survival: removes competing causes of death
- Relative survival
o Ratio of overall survival for cancer patients to the expected survival
of a comparable group of cancer-free individuals
o Provides a measure of excess mortality experienced by cancer
patients without requiring cause of death information
- Cause-specific (cancer-specific) survival
o Uses cancer death as the endpoint and censors people dying of
other causes
! Cancers is not the worldwide number one cause of death but it is the SECOND
=> first one are CVD’s!
2.2 Burden of cancer
- Since 2010: 26.3% increase in new cancer cases and 20.9 increase in
deaths
- Men: highest incidence prostate cancer
- Women: highest incidence breast cancer
- In Australia – melanoma cancer on the rise
- In the world – lung, trachea and bronchus on the rise because of the incline
in smoking
- The mean age of the world population is increasing rapidly -> that is why
the prevalence of cancer is also increasing => cancer is a disease of age!!
o If we standardise cancer for age => decrease!!
o Cancer is also not always linear increasing with age
- Increasing cancer incidence in adolescence and young adults
o Obesity, sedentary lifestyle, altered diet,
o Altered microbiome – AB use: hypothesis that everything is linked to
the gut microbiome
o Increased exposure to carcinogens
1
, o Physical inactivity has an effect on response of the therapy and
recovery after surgery
2.3 Causes of colorectal cancer
Risk factors: BMI, physical activity, smoking, alcohol, CRC family history, IBD, red
meat, fruit, vegetables
Colon cancer is not a single homogenous diseases
- Transcriptomic pathways: CMS1, CMS3, CMS2, CMS4
- (Epi)genomic: MSI and CIN with methylation, mutation or copy number
increase/decrease
- Driver genes: RAS and BRAF mutations
- Stroma-immune microenvironment
- The metastasis sometimes are showing a different CMS type then the
primarily tumor
3. Prevention – screening
3.1 Principles of cancer screening
The earlier you pick it up, the better the outcome!
Validity= the ability to predict who has the disease and who does not
- Sensitivity= the ability of a test to correctly identify those who have the
diseases
- Specificity= the ability of a test to correctly identify those who do not have
the diseases
Positive predictive value PPV= probability that a person with a positive test result
really has the disease
- Depends on test SE/SP but also on diseases prevalence
Likelihood ratio of a positive test result LR+
- Ptesting positive when having disease /Ptesting positive when not having disease
- When >1: test result is associated with disease
- Used to estimate post-test odds of disease based on performing a
diagnostic test
2
, Example: estimated prevalence of CRC in the general Belgian population is 0.6%.
The SE and SP of iFOB testing are 98%. What is the PPV of a positive test result (P
of having cancer) when 1000 persons are tested?
Bayes’ theorem= to calculate post-test probability when pre-test probability
differs from the prevalence in the reference group used to establish the positive
predictive value of the test
- Similar diagnostic test performance, the probability of disease will be very
different depending on the pre-test probability (usually prevalence)
- To calculate post-test probability:
1. Convert pre-test probability to odds: odds = P/(1-P)
2. Calculate likelihood ratio of (+) test results: LR+ = SE/(1-SP)
3. Calculate post-test odds: pre-test odds x LR+
4. Convert post-test odds to post-test probability: P = odds/(1+odds)
Example: Use of iFOB to detect colon cancer: SE = 70%, SP = 85%, 50y healthy
man: prevalence = 0.6%
In Flanders screening
- Screening for colon, breast and cervix cancer
- No screening for pancreas cancer which is really necessary
- Reasons why there is no screening for pancreas cancer: everything will
come out as false positive + there are no good screening programmes
Cancer screening
- Potential benefit: earlier diagnosis -> earlier stage cancer => improved
cancer specific survival
- Disadvantages:
o Overdiagnosis, overtreatment, complications related to diagnostic
procedure
o When low prevalence: even with high SP -> many false (+) (low PPV)
- Guideline for lung cancer
o Age 55-80
3
1. Cancer: epidemiology and risk factors
(Point) prevalence= proportion of existing cases (old and new) in a population at
a single point in time (cross-sectional)
Incidence= amount of new cases during a specified time interval (longitudinal)
- Risk= N cases/total population/t
- Odds= N cases/diseases free/persons over the study period => (risk/1-
risk)/ t
- Incidence rate= N cases/total person-years/t
2. Staging – TNM system
2.1 Measures of cancer survival and mortality
Overall survival: all cause, observed or crude survival => includes other causes
of death so it is an overestimation
Net survival: removes competing causes of death
- Relative survival
o Ratio of overall survival for cancer patients to the expected survival
of a comparable group of cancer-free individuals
o Provides a measure of excess mortality experienced by cancer
patients without requiring cause of death information
- Cause-specific (cancer-specific) survival
o Uses cancer death as the endpoint and censors people dying of
other causes
! Cancers is not the worldwide number one cause of death but it is the SECOND
=> first one are CVD’s!
2.2 Burden of cancer
- Since 2010: 26.3% increase in new cancer cases and 20.9 increase in
deaths
- Men: highest incidence prostate cancer
- Women: highest incidence breast cancer
- In Australia – melanoma cancer on the rise
- In the world – lung, trachea and bronchus on the rise because of the incline
in smoking
- The mean age of the world population is increasing rapidly -> that is why
the prevalence of cancer is also increasing => cancer is a disease of age!!
o If we standardise cancer for age => decrease!!
o Cancer is also not always linear increasing with age
- Increasing cancer incidence in adolescence and young adults
o Obesity, sedentary lifestyle, altered diet,
o Altered microbiome – AB use: hypothesis that everything is linked to
the gut microbiome
o Increased exposure to carcinogens
1
, o Physical inactivity has an effect on response of the therapy and
recovery after surgery
2.3 Causes of colorectal cancer
Risk factors: BMI, physical activity, smoking, alcohol, CRC family history, IBD, red
meat, fruit, vegetables
Colon cancer is not a single homogenous diseases
- Transcriptomic pathways: CMS1, CMS3, CMS2, CMS4
- (Epi)genomic: MSI and CIN with methylation, mutation or copy number
increase/decrease
- Driver genes: RAS and BRAF mutations
- Stroma-immune microenvironment
- The metastasis sometimes are showing a different CMS type then the
primarily tumor
3. Prevention – screening
3.1 Principles of cancer screening
The earlier you pick it up, the better the outcome!
Validity= the ability to predict who has the disease and who does not
- Sensitivity= the ability of a test to correctly identify those who have the
diseases
- Specificity= the ability of a test to correctly identify those who do not have
the diseases
Positive predictive value PPV= probability that a person with a positive test result
really has the disease
- Depends on test SE/SP but also on diseases prevalence
Likelihood ratio of a positive test result LR+
- Ptesting positive when having disease /Ptesting positive when not having disease
- When >1: test result is associated with disease
- Used to estimate post-test odds of disease based on performing a
diagnostic test
2
, Example: estimated prevalence of CRC in the general Belgian population is 0.6%.
The SE and SP of iFOB testing are 98%. What is the PPV of a positive test result (P
of having cancer) when 1000 persons are tested?
Bayes’ theorem= to calculate post-test probability when pre-test probability
differs from the prevalence in the reference group used to establish the positive
predictive value of the test
- Similar diagnostic test performance, the probability of disease will be very
different depending on the pre-test probability (usually prevalence)
- To calculate post-test probability:
1. Convert pre-test probability to odds: odds = P/(1-P)
2. Calculate likelihood ratio of (+) test results: LR+ = SE/(1-SP)
3. Calculate post-test odds: pre-test odds x LR+
4. Convert post-test odds to post-test probability: P = odds/(1+odds)
Example: Use of iFOB to detect colon cancer: SE = 70%, SP = 85%, 50y healthy
man: prevalence = 0.6%
In Flanders screening
- Screening for colon, breast and cervix cancer
- No screening for pancreas cancer which is really necessary
- Reasons why there is no screening for pancreas cancer: everything will
come out as false positive + there are no good screening programmes
Cancer screening
- Potential benefit: earlier diagnosis -> earlier stage cancer => improved
cancer specific survival
- Disadvantages:
o Overdiagnosis, overtreatment, complications related to diagnostic
procedure
o When low prevalence: even with high SP -> many false (+) (low PPV)
- Guideline for lung cancer
o Age 55-80
3
