The seasonality of pandemic influenza emergence

Spencer FoxPhD Student

Lauren Meyers LabThe University of Texas at Austin

spncrfx@gmail.com

@foxandtheflu

1889

1918

1957

1968

1977

2009

20

40

60

80

Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep

Influ

enza

Incid

ence

per

10,

000

Indi

vidua

lsPandemics emerge outside the typical

influenza season

1889

1918

1957

1968

1977

2009

Average US Seasonal Epidemic1997-2015

20

40

60

80

Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep

Influ

enza

Incid

ence

per

10,

000

Indi

vidua

ls

Fox et al PLoS Comp, 2017

Pandemics emerge outside the typical influenza season

1889

1918

1957

1968

1977

2009

Average US Seasonal Epidemic1997-2015

20

40

60

80

Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep

Influ

enza

Incid

ence

per

10,

000

Indi

vidua

lsShort-term, generalized immunity could prevent

emergence during seasonal epidemics

Fox et al PLoS Comp, 2017

1889

1918

1957

1968

1977

2009

Average US Seasonal Epidemic1997-2015

20

40

60

80

Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep

Influ

enza

Incid

ence

per

10,

000

Indi

vidua

lsShort-term, generalized immunity could prevent

emergence during seasonal epidemics

Generalized Immunity

Fox et al PLoS Comp, 2017

2009

0

50

100

150

200

Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep

Influ

enza

Incid

ence

per

10,

000

Indi

vidua

lsCombination of refractory period and seasonality

fit the historic pandemic emergence timing

Fox et al PLoS Comp, 2017

2009

0

50

100

150

200

Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep

Influ

enza

Incid

ence

per

10,

000

Indi

vidua

lsCombination of refractory period and seasonality

fit the historic pandemic emergence timing

Fox et al PLoS Comp, 2017

1. Two-strain model

2009

0

50

100

150

200

Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep

Influ

enza

Incid

ence

per

10,

000

Indi

vidua

ls

2008-2009

Combination of refractory period and seasonality fit the historic pandemic emergence timing

Fox et al PLoS Comp, 2017

1. Two-strain model2. Fit to 2008-2009 data

2009

0

50

100

150

200

Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep

Influ

enza

Incid

ence

per

10,

000

Indi

vidua

ls

2008-2009

Combination of refractory period and seasonality fit the historic pandemic emergence timing

Fox et al PLoS Comp, 2017

1. Two-strain model2. Fit to 2008-2009 data3. Test when pandemics emerge

2009

0

50

100

150

200

Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep

Influ

enza

Incid

ence

per

10,

000

Indi

vidua

ls

2009

0

50

100

150

200

Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep

Influ

enza

Incid

ence

per

10,

000

Indi

vidua

ls

2008-2009

Combination of refractory period and seasonality fit the historic pandemic emergence timing

Fox et al PLoS Comp, 2017

1. Two-strain model2. Fit to 2008-2009 data3. Test when pandemics emerge

Pandemics that are successful emerge “slower” within the refractory period0.0

0.1

0.2

0.3

Jan Feb Mar Apr May Jun Jul AugPan

dem

ic E

mer

genc

e P

roba

bilit

y

2.0

2.2

2.4

Jan Feb Mar Apr May Jun Jul Aug

Pan

dem

ic R

eff

Fox et al PLoS Comp, 2017

Pandemics that are successful emerge “slower” within the refractory period0.0

0.1

0.2

0.3

Jan Feb Mar Apr May Jun Jul AugPan

dem

ic E

mer

genc

e P

roba

bilit

y

2.0

2.2

2.4

Jan Feb Mar Apr May Jun Jul Aug

Pan

dem

ic R

eff

Fox et al PLoS Comp, 2017

Pandemics that are successful emerge “slower” within the refractory period0.0

0.1

0.2

0.3

Jan Feb Mar Apr May Jun Jul AugPan

dem

ic E

mer

genc

e P

roba

bilit

y

2.0

2.2

2.4

Jan Feb Mar Apr May Jun Jul Aug

Pan

dem

ic R

eff

Fox et al PLoS Comp, 2017

Conclusions

2009

0

50

100

150

200

Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep

Influ

enza

Incid

ence

per

10,

000

Indi

vidua

ls

Conclusions•Reasonable

•Still need better understanding of short-term immunity, but it could explain pandemic emergence seasonality

2009

0

50

100

150

200

Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep

Influ

enza

Incid

ence

per

10,

000

Indi

vidua

ls

Conclusions•Reasonable

•Still need better understanding of short-term immunity, but it could explain pandemic emergence seasonality

•Handwavy, but (maybe) cool implications

•Optimizing pandemic surveillance systems and response

•Seasonal flu vaccination2009

0

50

100

150

200

Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep

Influ

enza

Incid

ence

per

10,

000

Indi

vidua

ls

Pandemic wave dynamics

0

100

200

300

Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

incidence

Pandemic wave dynamics

0

100

200

300

Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

incidence 2009 seasonal

epidemic pandemic 1st wave

pandemic 2nd wave

Pandemic wave dynamics

0

100

200

300

Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

incidence 2009 seasonal

epidemic pandemic 1st wave

pandemic 2nd wave

Cities with larger seasonal flu epidemics had smaller initial pandemic waves

0

1

2

3

4

5

1 2 3 4Seasonal Peak Intensity

1st W

ave

Pdm

Pea

k In

tens

ityCutoff=1e+06

Cities with larger seasonal flu epidemics had smaller initial pandemic waves

0

1

2

3

4

5

1 2 3 4Seasonal Peak Intensity

1st W

ave

Pdm

Pea

k In

tens

ityCutoff=1e+06

Many other factors to investigate, and needs some more intricate modeling

0

1

2

3

4

5

1 2 3 4Seasonal Peak Intensity

1st W

ave

Pdm

Pea

k In

tens

ityCutoff=1e+06

Many other factors to investigate, and needs some more intricate modeling

0

1

2

3

4

5

1 2 3 4Seasonal Peak Intensity

1st W

ave

Pdm

Pea

k In

tens

ityCutoff=1e+06

1. Nonlinear effects

Many other factors to investigate, and needs some more intricate modeling

0

1

2

3

4

5

1 2 3 4Seasonal Peak Intensity

1st W

ave

Pdm

Pea

k In

tens

ityCutoff=1e+06

1. Nonlinear effects2. Spatial clustering

Many other factors to investigate, and needs some more intricate modeling

0

1

2

3

4

5

1 2 3 4Seasonal Peak Intensity

1st W

ave

Pdm

Pea

k In

tens

ityCutoff=1e+06

1. Nonlinear effects2. Spatial clustering3. Climatic factors

Acknowledgements• Lauren Ancel Meyers

• Joel C. Miller

• Cécile Viboud

• Funding

• NIH MIDAS program

• DTRA

• National Geographic Young Explorer Program

spncrfx@gmail.com

@foxandtheflu

Spencer’s contact info:

City flu data may hold clues

San Bernardin San Diego Santa Ana St. Louis

Los Angeles Oakland Philadelphia Phoenix

Atlanta Chicago Denver Detroit

Jan Apr Jul Jan Apr Jul Jan Apr Jul Jan Apr Jul

0.6

0.9

1.2

1.5

0.5

1.0

1.5

2.0

2.5

0.50

0.75

1.00

1.25

0.4

0.8

1.2

1.6

0.5

1.0

1.5

0.50

0.75

1.00

1.25

1.50

0.5

1.0

1.5

2.0

0.6

0.9

1.2

1.5

0.5

1.0

1.5

2.0

0.4

0.8

1.2

1.6

0.75

1.00

1.25

1.50

1.75

0.5

1.0

1.5

2.0

ILI %

Seasonal epidemics and short-term immunity create pandemic “refractory period”

0

50

100

Jan Feb Mar Apr May Jun Jul Aug

Seas

onal

Inci

denc

e

0.0

0.1

0.2

0.3

Jan Feb Mar Apr May Jun Jul Aug

Pand

emic

Em

erge

nce

Prob

abilit

y

Fox et al PLoS Comp, 2017

Seasonal epidemics and short-term immunity create pandemic “refractory period”

0

50

100

Jan Feb Mar Apr May Jun Jul Aug

Seas

onal

Inci

denc

e

0.0

0.1

0.2

0.3

Jan Feb Mar Apr May Jun Jul Aug

Pand

emic

Em

erge

nce

Prob

abilit

y

Fox et al PLoS Comp, 2017

Seasonal epidemics and short-term immunity create pandemic “refractory period”

0

50

100

Jan Feb Mar Apr May Jun Jul Aug

Seas

onal

Inci

denc

e

0.0

0.1

0.2

0.3

Jan Feb Mar Apr May Jun Jul Aug

Pand

emic

Em

erge

nce

Prob

abilit

y

Fox et al PLoS Comp, 2017

Seasonal epidemics and short-term immunity create pandemic “refractory period”

0

50

100

Jan Feb Mar Apr May Jun Jul Aug

Seas

onal

Inci

denc

e

0.0

0.1

0.2

0.3

Jan Feb Mar Apr May Jun Jul Aug

Pand

emic

Em

erge

nce

Prob

abilit

y

Fox et al PLoS Comp, 2017

Influenza pandemics are defined by the two main antigens

Influenza pandemics are defined by the two main antigens

How can there be strain replacement following pandemics

How can there be strain replacement following pandemics

Immunological Duration (Years)

Homosubtypic Immunity

0 0.1 1 10 100

H1N1 H3N2

Increased Susceptibility

Decreased Susceptibility

Immunological dynamics underlying flu spread

Immunological Duration (Years)

Homosubtypic Immunity

0 0.1 1 10 100

H1N1 H3N2

Increased Susceptibility

Decreased Susceptibility

Immunological dynamics underlying flu spread

Immunological Duration (Years)

Homosubtypic Immunity

0 0.1 1 10 100

H1N1 H3N2

Increased Susceptibility

Decreased Susceptibility

Immunological dynamics underlying flu spread

Immunological Duration (Years)

Long-term HSI

Homosubtypic Immunity

0 0.1 1 10 100

H1N1 H3N2

Increased Susceptibility

Decreased Susceptibility

Immunological dynamics underlying flu spread

Immunological Duration (Years)

Short-term HSI

Long-term HSI

Homosubtypic Immunity

0 0.1 1 10 100

H1N1 H3N2

Increased Susceptibility

Decreased Susceptibility

Immunological dynamics underlying flu spread

Recent experiments show example of short-term HSI

Recent experiments show example of short-term HSI

Recent experiments show example of short-term HSI

Influenza pandemics consistently top the list of most worrisome infectious diseases