20.102 | Spring 2005 | Undergraduate
Macroepidemiology (BE.102)

Lecture Notes

The following table provides an outline of each lecture.

lEc # TOPICS
1

Introduction: Dimensions of information available to define the unknown causes of common diseases.

1. Four essential forms of information:

(a) Public health records: historical changes in age-specific mortality and/or incidence. 
(b) Environmental epidemiology and community risk. 
(c) Population genetics and familial risk. 
(d) Human physiology and genetics.

2. Common theoretical structures:

(a) Clonal expansion models: carcinogenesis and atherogenesis 
(b) Cell/function mortality models: diabetes, neural degenerative diseases

3. Discussion of expectations and diseases of interest to class members.

2

Public health records and estimation of age-specific risk.

1. Introduction to birth cohort and age specific mortality and incidence data sets (see Mortality Data).

(a) Accuracy of the public record for mortality (MIT) and incidence (SEER) data 
(b) Accuracy of the U.S. census data. 
(c) Uncertainty in calculating an age-specific mortality or incidence rate: OBS (h,t) 
(d) Historical changes in diagnosis, prevention and therapy: S (h,t) 
(e) Historical changes in accuracy of recording causes of death: R (h,t) 
(f) Consideration of competing forms of death within an age interval: TOT (h,t) 
(g) Best (?) present estimate of age-specific expectation of risk for any disease: OBS (h,t)/{R (h,t) x [1-S (h,t)] x [1-TOT (h,t)]}

3

Effects of changes in diagnosis, prevention and therapy on historical mortality rates. Sources of differences in reported incidence and mortality.

1. Early diagnoses: cervical, breast and prostate cancers.

2. Prevention: food, water and medical microbiology hygiene; smoking avoidance.

3. Therapy: antibiotics and infectious diseases. Chemotherapy and skin cancers, pediatric leukemia,  Hodgkin’s lymphoma and adult chronic myelocytic leukemia.

4

Mathematical models derived from histopathological observations and age-specific mortality/incidence data. Examples: lung and colorectal cancers.

1. Histology and histopathology of normal epithelium, preneoplasia and neoplasia: stemcells identified (?)

2. The Armitage-Doll two (rate-limiting) stage model of carcinogenesis

3. Genetic changes in carcinogenesis. (Some tested and untested hypotheses.)

(a) Initiation 
(b) Promotion: required number of events “m”, growth of preneoplasia, effect of stochastic factors on expected numbers of “cells at risk”, rates of required genetic changes (if any). 
(c) Progression: invasion and metastases. Simplifying assumptions.

5

Initiation

1. Required number of events, “n”.

2. Growth of tissues.

3. Concept of maintenance turnover units.

4. Definition of “cells at risk”.

5. Rates of required genetic changes.

6. Definition of Cinit(n).

6

Promotion and Progression

1. Promotion

(a) Required number of events, “m”. 
(b) Growth of number of cells at risk. Definition µ = (α-β). 
(c) Continuation of turnover units. 
(d) Rate of hypothetical required events. 
(e) Definition of Cprom(m).

2. Progression: Heterogeneity and general confusion of required genetic changes.

3. Demonstration of CancerFit Program.

7

Multiparametric Analysis of Colon Cancer

1. Somatic genetics of colon carcinogenesis. Assumption of n=2.

2. Range of parameters without limitation.

3. Limitations on parameters.

(a) Minimum estimate of F ~ fraction of population with preneoplastic lesions, f = 1? 
(b) Absence of new polyps after age 60. Initiation limited to juvenile period? 
(c) Application of mutation and loss of heterozygosity levels in human organs. 
(d) Assumption that only stem cells are at risk.

4. Re-estimation of parameters using limitations derived from observations.

8

Multiparametric Analysis of Lung Cancer

1. Somatic genetics of lung carcinogenesis. Assumption of n=2.

2. Range of parameters without limitation.

3. Limitations on parameters.

(a) Epidemiology of lung cancer: Rise of F with E, F=E? 
(b) Epidemiology of early death among smokers, f~1/6? 
(c) Fraction of smoking population with preneoplastic lesions. 
(d) Initiation limited to juvenile period? 
(e) Application of mutation and loss of heterozygosity levels in human organs. 
(f) Assumption that only stem cells are at risk.

4. Re-estimation of parameters using limitations derived from observations.

5. Testing of derived model among persons who quit smoking.

9

Role of Gender

1. Growth periods

2. Total stem cell number and/or gender-specific physiology/lifestyles.

10

Sub-populations at Risk

1. Independence and/or intersections of inherited and environmental risks.

2. Essential and accelerating risks.

3. Stochastic risk.

11

National Risk and Community Risk

1. National Risk

(a) Comparative history of cancer rates among European, African Americans and Japanese. 
(b) Hypothetical role of post-agrarian environment versus genetic risk differences.

2. Community Risk: Risk factors in Pennsylvania (1958-1995)

(a) Use of binomial distribution to define expected age-specific mortality rates for any community in a state given the number of persons at risk and the average age specific mortality rate for the set of all communities in the state: null hypothesis that communities have identical cancer mortality rates. 
(b) Use of Kolmogorov-Smirnov analysis to discover significant differences between observed and expected distributions of mortality rates distributed over all communities.

12

Familial Risk Expectations

1. Familial environmental risks: cultural and idiosyncratic habits

2. Familial genetic risks: homozygous, heterozygous, nullizygous and functionally dominant risks.

3. Siblings, dizygotic and monozygotic twins.

4. Spouses.

5. Parents and children.

6. Pleiogenic risks.

13

Familial Risk Observations

1. Parents and children

2. Spouses

3. Twins

4. Pleiogenic risks

14

Population Genetics

1. Age and size of human population.

2. Rates of mutation.

3. Rates of population growth and survival of neutral alleles.

4. Dominant, recessive and non-deleterious conditions.

5. Hardy-Weinberg equilibrium and expectations of allele frequencies.

15

Population Genetics (cont.)

1. Allelic distributions and expected frequencies (see the Human Gene Mutation Database)

2. Generation of multiallelic risk by gametogenic mutations: mutational spectra.

3. Expectations for deleterious mutational spectra.

4. Expectations for non-deleterious spectra.

5. Examples of multiallelic risk for deleterious conditions (~2000).

6. Examples of mono-allelic risk for deleterious conditions (sickle cell anemia and cystic fibrosis)

7. Example of multi-allelic risk for a common disease (MC1R and risk of skin cancer).

16

The search for genes carrying mutations conferring risk for common diseases.

1. The assumption of mono-allelic risk and the resulting “great SNP hunt”.

2. The assumption of multi-allelic risk and the probable need for pangenomic scanning.

17

The search for genes carrying mutations conferring risk for common diseases. (cont.)

1. Technology for DNA-based population genetic searches: multi-capillary DNA sequencers, SNP detectors and multi-capillary denaturing electrophoresis for multiple allele discovery.

2. Costs and throughput: what is feasible and fundable?

3. Strategy for pangenomic search for 100 common diseases: instrumentation, DNA samples and general statistical model.

18

The search for genes carrying mutations conferring risk for common diseases. (cont.)

1. “Genicity” of risk.

2. “Zygosity” of risk.

3. “Ethnic stratification” and problem of mixedness.

4. Potential use of “age-specific allelic decline” to identify genes carrying risk for mortal diseases.

19

The search for genes carrying mutations conferring risk for common diseases. (cont.)

1. Using familial and general population risks for colon cancer to limit hypotheses.

20 The search for genes in which mutations of initiation (and promotion, if any) occur in common human cancers.
21

The origins of somatic and inherited mutations in humans.

1. The environmental mutagen hypothesis.

2. The search for environmental mutagens.

3. The attempts to match in vitro human cell spectra with in vivo human tissue spectra.

4. The discovery of dose-dependence of chemically induced mutational spectra.

5. The discovery of shared subspectrum of human mutations with DNA methylating mutagens.

22

The origins of somatic and inherited mutations in humans. (cont.)

1. Mutations of mitochondria and nuclear genes in smokers and non-smokers’ lungs.

2. Mutations of mitochondrial and nuclear genes by DNA polymerases.

3. Possible restriction of somatic stem cell genetic changes to juvenile years: chromosomal translocations and point mutations as a function of age.

23

Hypotheses that need testing. (Professor turns 60 and looks to future.)

1. Point mutation and chromosomal translocation mutations that initiate preneoplasia are restricted to years prior to metamorphosis to adult form in humans.

2. DNA turnover in human stem and transitional cells involves a significant fraction of DNA each day.

3. Point mutations in tissues are primarily generated by DNA polymerase errors during DNA synthesis and re-synthesis during DNA turnover in pre-adult stem cells.

4. Environmental risk factors accelerate growth of preneoplastic lesions that arose in juvenile years.

5. Environmental risk factors induce organ-specific inflammation.

6. Genetic risks for common diseases are multi-allelic and commonly monogenic.

24

Evolution, embryogenesis and carcinogenesis. (Guest lecturer: Dr. Elena V. Gostjeva)

1. Novel histologic preparation of tissues and high resolution microscopic images.

2. Human embryogenesis: discovery of multiple nuclear morphotypes and “metakaryotic” modes of nuclear division.

3. Human embryogenesis: discovery of asymmetric metakaryotic nuclear divisions: evidence of organogenic stem cell behavior.

4. Human carcinogenesis: normal, preneoplastic and neoplastic histology.

5. Human carcinogenesis: persistence of juvenile phenotype and transformation to fetal phenotype.

6. Potential for high risk of genetic change in metakaryotic DNA synthesis and chromosomal segregation.

7. Evolutionary origin of metakaryotic phenotype.

Putting it together: Gostjeva-Thilly cascade model for risk and genesis of common clonal diseases.

25-27

General discussion and student presentations.

Students are encouraged to invite academic advisors and friends to presentations.

Course Info
As Taught In
Spring 2005