Loading document…
Opening in Pages for Mac...
Your browser isn’t fully supported.
For the best Pages for iCloud experience, use a supported browser.
Learn More
Cancel
Continue
Study
Pop
Class
Design
Results
Meas'd
👶
/
👩
🎓
?
👶
/
👩
🎓
high
er?
Meas'd
🐾
?
🐾
Y/N?
Meas'd
👿
?
👿
Y/N?
Yewdell/
Santos
Francis et
al. 1947
94
👶
Before-and-after HI titers following "A prime" (Rhodes)
infection and hospitalization (time of sampling not given)
after 1946/7 PR8+Weiss ("A" H1N1) +B vaccine program.
See right and
note.
👶
post-
variant
inf
.
Y absolute
values; N fold
increase.
X
--
X
--
Davenport
et al. 1953
(Francis 1)
--
👶
Pooled sera collected pre/post 1947, or in 1952, compared
for anti-strain titers. No actual individual exposure histories
measured (pooling erases individual gains to later strains.)
"
👶
Abs highest"
👶
(pooled
)
Y (w/ 1949
strain instead
of a 1947 one)
X
--
X
--
--
Davenport
Hennessy
1955
100s
👶
👿
A 1953 FM1 vaccine trial, this time with PR8-imprinted
recruits who would have had 6 years of exposure to FM1
and related ("A prime") H1N1 strains. This is the perfect
(unintentional) OAS experiment since subjects have
exclusively been exposed to two H1N1 variants in lifetime.
👶
Abs near-equal
before; lower after
vax (very little
back-boosting)
.
👶
(pooled
)
N before vax
(near-equal) or
after vax
(lower)
X
--
👿
N:
Subjects were born in 1935 or earlier, so PR8 should "orient" the
natural response to FM1-strains between 1947-53 and to the FM1 vax.
Instead, PR8 back-boosting after vax is minimal. This suggests novel,
pre-primed FM1 "imprinting" to a
non-childhood H1N1 variant
.
Symptomatic flu is lower in vaxxed vs. control (83
%
effi cacy).
--
Davenport
et al. 1955
(Francis
1.1)
3450
👶
Repeat of "Francis 1" with comparison of new sera pools
collected in Michigan and Sheffi eld, England in 1953.
Importantly, a 1947 strain is selected to represent "A
prime" (so that prototype strains are used for all models).
High abs for "A
Prime" in ages
that looked low in
prev. study.
👶
(pooled
)
Occasionally.
All groups are
high for new A
Prime strain.
X
--
X
--
--
Jensen
et al. 1956
(Francis 2)
100s
👶🐾
👿
Donors are grouped by age to infer imprinting strain (SW,
PR8, or FM1). Serum-absorption before and after
vaccination with SW, PR8, or FM1 in each group.
👶
Abs, imprint-
ing inconsistent
after vax
.
👶
(pooled
)
N after vax in
5/6 groups
.
🐾
only
in 4/9
group
👿
N: imprinting is only semi-common. Thus, all age groups can make
antibodies specifi c to non-childhood strains despite being "sinned."
--
Davenport
Hennessy
1957
100s
👶
👿
Further results from 1956 donor set. Donors in 3 age
groups given non/adjuvanted SW, PR8, FM1, or FM1-
related vaccines. Children then vaxxed w/ new strain a year
later, indirectly measuring for any detriment from prior vax.
👶
Abs inconsis-
tent after vax. Abs
higher in cross-
vaxxed kids
.
👶
(pooled
)
N after vax in
3/6 groups
(Fig 1).
X
--
👿
N: In children given "Malaya" vax, 6-week abs are 3x higher for those
previously cross-strain-vaxxed than unvaxxed controls (Table 2). (So,
authors cite "persistence" (as in literal non-disappearance) of prior
response as evidence for "orienting" of future response.) Double poly-v
vax leads to full spectrum immune response in kids and adults.
--
MRC,
1958
1270
👿
Anti-"Asian" (H2N2) vaccine in H1N1-imprinted teen boys
in 1958, vs. controls receiving tri-H1N1 or B vaccine.
H2N2-specifi c
immunity leads to
vax effi cacy vs
both controls.
X (not
antibod-
ies)
--
X
--
👿
(sort-of)
N: Technically cannot measure negative effect from H1N1 as no true
naive controls, and no measure of back-boosting effect. However, H2N2
vax effi cacy demonstrates novel immunity vs. a new subgroup despite
imprinting. (Francis retconned H2N2 into OAS in 1960.)
--
Francis,
1960
("Doc-
trine")
--
🤦
♂
️
Restatement of Davenport et al. 1953 results and gross
misrepresentation of Jensen et al. 1956 results, plus some
theorizing.
There are no true
"results." This
paper should
never be cited.
X
--
X
--
X
--
--
De St.
Groth/
Webster,
1966
58
👶🐾
👿
Pre-and-post SW-vax sera from FM1-imprinted kids in
1954 (collected 14 and 42 days after) or 1955 (collected at
42 days); absorption and avidity (equilibrium constants, K)
used to assess specifi city.
note
👶
Y, but less
than 10x
🐾
Y, but
not
100
%
👿
N: post-vaccine SW antibodies are less avid vs. SW than FM1 but are
more
avid vs. SW than an "expected" primary response (the entire logic
quoted by authors to conclude imprinting). SW specifi c antibodies are
generated.
--
Webster,
et al. 1976
50
👩
🎓
🐾👿
In (presumably H2N2-childhood-strain) young adults, HI
titers against target (Port Clalmers73) and prototype (HK68)
H3N2 strains compared before and after admin. of whole
and subunit PC vaccines (incl. 2 bivalent with a B strain)
note
👩
🎓
HK=
likely 1st
H3
Only in 5/10
groups,
& <10x (Most
others are =)
🐾
some
times
👿
N:
Higher
anti-PC
titers
in HK-"sinned" after vaccination with PC
(2,000-2,200)
, vs. not-
HK-
"sinned"
(480)
(Table 4).
--
Nobel, et
al. 1977
475
👶
"Back-boost" responses 21 days after receipt of 1976
"swine" H1N1 vaccine (NJ/76), by presumed childhood
strain (PR8, FM1, H2N2, H3N2). Additional results (with 509
different subjects) presented for swine-cross-reaction after
H3N2 infection and vaccination.
Back-boost
limited to in-
subgroup (H1N1
no boost H2/3N2
and vice versa)
👶
N before and
after vax. for
middle-aged
adults. See
note.
X
--
👿
N: Anti-NJ/76 GMTs in H1N1-imprinted is than younger H2N2/H3N2
groups, though later draws would probably show higher results in
younger set; and since H2N2/H3N3 fold changes are sub 2X there is no
"anamnestic" response to blame. Vs.-H1N1-back-boost GMTs are
comparable (123 vs 167 in FM1) or lower (118 vs 54 in PR8) (Table 2)
Hoskins et
al. 1979
(of
"paradox"
fame)
100s
🤦
♂
️
📦
👿
Infection outcomes in teens observed in three variant H3N2
waves (1972,74,76) preceded by vax with pre-variant strain
(1972 & 76) or variant strain (1974) vs. control vax (B strain).
However, "repeat vax" is only shown to underperform new
vax in 1974; no "new vax" group exists in 1976 to compare
with. Most data does not support "paradox."
All groups
gradually reach
~50
%
conf.
infected rate.
X
--
X
--
👿
N
for inf.
-
prior H3N2 infection highly protective
against variant strains
in
three successive waves.
N for vax: Performance of "repeat vax" group is
not consistently worse than "fi rst vax;" 1976 offers no "fi rst vax" group to
compare with (Beyer, WEP et al. 1998); variant strains and antigen
mismatch do not appear to influence effi cacy in any way. As with
Francis's "Doctrine," this is a case of misrepresented results.
--
Masurel et
al. 1981
378
👩
🎓
👿
Sequential H1N1 vaccinations in schoolboys: 2-course
"Russia" responses compared in previously "NJ" (swine)-
vaccinated vs. not. Draws are 4 weeks after 1st Russia
vaccine and 3 weeks after 2nd.
NJ-vaxxed have
diminished Russia
response / high
NJ-backboost
👩
🎓
Y, especially
after 2nd
Russia vax.
X
--
👿
Y: NJ-vaccinated kids show
lower
response
vs controls
to Russia after
2nd
Russia vax
. Escalating NJ-backboost in both cases suggests strong
antigenic masking. Again reinforces Hoskins, et al. results and shows
effect in a
non-childhood
-strain.
Thus, antigenic interference causes
repeat vaccination failure for in-subgroup variants.
--
Beyer et
al. 1986
50
👶🐾
👿
Day-28 antibody rises after trivalent (H1N1+H3N2+B)
vaccine are compared by age (born before/after 1950)
See right. Most
data too
processed /
obscured for fi ne
interpretation.
👶
Is not shown.
🐾
some
times
👿
Results are heavily obscured by various means. However, apparent Y for
IgG/A (more nonresponders in non-primed seronegative groups, Table 2)
and N for IgM, (non-primed seronegative had high IgM responses, Fig 2),
demonstrating novel (but still developing) immune response.
Better if
more time before draw.
--
Powers
Belsche
1994
?
👶
👿
Responses to contemporary H1N1 vs. older H1N1
evaluated in three age groups after 1991 trivalent vaccine.
(From abstract)
Older groups
have contem-
porary Abs.
👶
N
X
--
👿
N - Authors conclude immune senescence better accounts for pre- and
post-vaccine antibody dynamics. (I didn't purchase this one, since it
doesn't claim to support OAS.)
--
Keitel, et
al. 1997
4003
📦👿
Hoskins paradox in adults? Vaccine effi cacy in fi ve
consecutive yearly randomized trials, with returning,
previously vaxxed subjects being re-vaxxed up to 6X.
No Hoskins.
X
--
X
--
👿
N - Repeat-vaccinated adults fare just as well as fi rst-timers. Anti-
vaccine-strain antibody responses are only slightly diminished in repeat-
vaxxed, and effect does not increase after 3rd vax. I would note that
predicting long term, future maintenance of these results requires
assuming no eventual detriment from immune debt, tolerance, etc.
--
Wrammert
et al. 2008
10
🐾
Responses after trivalent vaccination.
Focus is on
plasma
cells and
influenza
B
vax response;
a bit
left-fi eld.
note
X
--
🐾
Y & N
X
--
No OAS
Moody et
al. 2011
11
👿
Trivalent and infection challenge responses are evaluated
for rises in non-specifi c flu antibodies. Since childhood
strain is not considered, this is not OAS. However,
challenge responses provide data RE: detriment?
note
X
--
x
(back-
wards
design)
--
👿
(indirect
meas.)
N: worst symptom-haver antibody increases are not different than others
in challenge (Fig S1)
OAS lol
Wrammert
et al. 2011
(Wram. 2)
9
🐾
👿
Plasma cell / antibody dynamics
~
10 days
after infection
with 2009 "swine" flu (pH1N1). Once again too plasma-cell/
short-term-response-focused for an "OAS" paper.
However, some fi ndings fi t in this table.
63
%
,
i.e.
not all,
plasma cells are
non
-
pH1N1
specifi c
X
--
🐾
63
%
active
Bcells
(early)
👿
(indirect
meas.)
N
:
pH1N1 infection generates 1/3 specifi c (but lower-avidity) and 2/3
cross-reactive, high-avidity antibodies (Fig 2). Most severe infection
is
in
patient
(30, female)
w/
least-mutated (not-primed)
H1N1 antibodies
(Fig
S3)
.
Prior immunity thus protective, does not forbid new immunity.
OAS
Stem Abs
Esposito
et al. 2011
69
👶
👿
Sera collected ~28 days after pH1N1 infection in children.
HI seropositivity and titers compared in children w/ pre-
existing antibodies for human H1N1.
See right.
👶
N: 96
%
sero-
conversion in
"sinned."
Titers similar.
X
--
👿
N: Prior immunity to H1N1 does not prevent seroconversion after pH1N1
infection. Arguably, higher severe rates suggest a detriment; however,
ADE is just as plausible, and study design is biased toward rare severe
outcomes to begin with (
all
subjects presenting to ER
1st
).
--
Carter et
al. 2012
846
👶
Sera collected from individuals aged born years 1920-2010
and tested for HI against pH1N1, 1918 sequence-based
HA, and historical H1N1 and H3N2 strains.
Older groups defy
doctrine. Plus,
evidence of
antigenic rever-
sion in H1N1.
👶
N: Older
groups defy
doctrine.
Examples in
note.
X
--
X
--
--
Lessler et
al. 2012
151
👶
"OAS" in historical H3N2 strains? 151 donors between 7
and 81 years old, with low vaccination rates (66
%
never
vaccinated; + 13
%
not in last 5 years), tested for
neutralization of various prototype H3N2 models.
No OAS in H3N2.
Adults neut'lz old
& new strains
equally.
👶
N in adults
(see note)
X
--
X
--
--
Li et al.
2012
(Wram. 3)
24
🐾
👿
Plasma cell / B Cell / antibody dynamics 28 days after
pH1N1 subunit vaccine in 24 adults 26-64 y old (flu-
positive b cell isolation and mAb characterization).
Early, high
expansion of
H1N1-HA-cross-
reactive B Cell
memory
X
--
🐾
Y, but
not
100
%
👿
(indirect
meas.)
N: Imprinting enables a rapid remodeling of the B Cell pool to prioritize
neutralizing cross-reactive antibodies (bind to both the vaccine and
previous circulating H1N1 strains), including neutralizing stem-binding
antibodies. Also, some vaccine-specifi c antibodies appear (Fig 3.)
(Possibly more if more time before draw.))
OAS
Stem Abs
Study
Pop
Class
Design
Results
Meas'd
👶
/
👩
🎓
?
👶
/
👩
🎓
high
er?
Meas'd
🐾
?
🐾
Y/N?
Meas'd
👿
?
👿
Y/N?
Yewdell/
Santos
Li et al.
2013
(Wram. 4)
54
👶🐾
👿
Sera collected ~10 days (adults) or ~28 days (children) after
pH1N1 compared for HI-dependence on K133 epitope,
present in pH1N1 and in human H1N1s from 1983 to 1996.
See note.
Questionable due
to potential early
sera collection.
👶
Y for K133
epitope.
🐾
Y
(early
draw)
👿
N: Immune experience with K133 arguably "limits" (early) reply to other
epitopes after pH1N1 infection, but only because pH1N1
has K133
- and,
since naive infection in children leads to K133 antigens (table S3 subjects
25, 30, 46-54), the "OAS" response
replicates
a naive response.
Obviously there is no clear "detriment" to recall response to retro/
reversion antigens (likely
protective
in older age).
OAS
Ellebedy
et al. 2014
(Wram. 5)
78
🐾
H5N1 vaccine responses evaluated before and 28, 180,
and 208 days after fi rst dose, with booster after day 180.
D28 only shows
X-reactive stem
Abs; booster
reveals abundant
novel HA1 Abs.
X
--
🐾
N for
long
term
respo
nse
X
--
OAS
Stem Abs
O'Donnell
et al. 2014
58
👶
👿
An actual, rational design to validate OAS
. HI
for
pH1N1 vs.
childhood strains compared
before and 28 days after 2010
pH1N1 trial (previously conf. infected excluded)
.
Unfortunately, raw titers are not given for childhood strains.
Extreme lack of
back-boost after
pH1N1 vax.
👶
(post-
vax fold
change)
N for post-vax
"back-boost"
for all 3
"H1N1 eras"
X
--
👿
N:
Huge
N. Hindenburg blimp jpg
-level
N. All
3
age groups mount novel
pH1N1-responses
(fold change from previous)
with little back-boost.
No OAS
Andrews
et al.
2015a
(Wram. 6)
21+4
👶🐾
👿
pH1N1 fi rst or second-time (6-12months) vax responses
evaluated at 0, 5-7 (plasmablasts), and 14-21 days
(memory b cells) <- obviously focusing on recall responses.
(Flu-positive b cell isolation and mAb characterization.)
Memory B Cell
expansion to
conserved stalk
epitopes is 1-off.
Head abs less
auto-reactive.
👶
N: pre-and-
post-2nd-vax
abs are
pH1N1 spe-
cifi c, resemble
non-childhood
immunity
🐾
N:
(note)
👿
N: While fi rst-time pH1N1-vax recipients with low preexisting titers (Fig
5B) react with broadly cross-reactive stem abs (Fig 3B), second time
recipients almost all have high preexisting titers, no stem ab response,
and more pH1N1-specifi c response (less cross-reactive; Fig 4E).
Additionally, stem abs are shown to be less whole-virus binding (Fig 6)
and more auto-reactive (Fig 7) implying detriment to long term stem
recall/benefi t to the long-term generation/retention of new anti-head abs.
OAS
observed
after fi rst
pH1N1
exposure
(lol)
Gostic et
al. 2016
1515
👶
👿
Hospitalizations for avian flus are cross-referenced with
birth year flu subgroups, to see if childhood H1 exposure
protects against sibling species H5 and H3 against H7.
Model predicts
outcomes.
👶
(Protect-
ion from
cases.
)
Y
X
--
👿
Y/N: There is no "met H3 in adulthood without prior H1 exposure" control
group to demonstrate that lower H3 protection results from H1 immune
interference as opposed to less overall exposure, immune senescence,
etc. Extreme demographic/exposure confounders (all H7 cases in one
country; low sample size) are also a hazard.
--
Gostic et
al. 2019
9451
👶
👿
In a hospitalization-heavy data set, PCR-identifi ed H1N1
and H3N2 cases between 1993 - 2015 are cross-
referenced to birth year subgroups.
Cases lower than
trend to
👶
sub-
group except
when not, i.e. for
1945-56 vs H3N2
.
👶
(Protect-
ion from
cases.
)
A model
incorporating
childhood to
young adult
exp'r likely
outperforms.
X
--
👿
N as regards "childhood strain." Since individuals born
~1945 still have
protection against H3N2 care (Fig 3D; neither the HA nor NA 1st-
exposure-model correctly predicts when protection ends), they
developed lifelong immune memory to a post-childhood strain in their
teens. Lack of H3N2 protection in individuals born before 1945 is thus
immune senescence
(they met H3N2 as adults), not OAS.
--
Study
Pop
Class
Design
Results
Meas'd
👶
/
👩
🎓
?
👶
/
👩
🎓
high
er?
Meas'd
🐾
?
🐾
Y/N?
Meas'd
👿
?
👿
Y/N?
Yewdell/
Santos
