What constitutes a simple, reliable, message for outcome in a rare disease?
Cystic Fibrosis in Europe
Anil Mehta FRCP FRCPCHUniversity of Dundee, Scotland
An argument for governments from the ECFRG
For full details see: www.lancet.com (header section: audio; march20)
1. Historical Context podcast
2. European CF3. EU versus non EU (2003)
50,000 years ago0−2Κ
∆F508, F508del or Phe508del
We are here
2010
-
10K10K
−10Κ
3 base pairs lost on chromosome 7 in a single gene cftr
Imagine the first modern human carrier of CF
Fibrogenicum (- us or - a)
-
-
---- - --- ---- -- - -- - -- -- --
Population50,000 years ago0−2Κ
ΑD
∆F508
We are survivors
2010
BCBC
-
10K10K
−10Κ
BUT: 3 in every 4 children in recent history died before their 6th birthday
75% died before 6 years of age
Genetically fit
Fibrogenicum super-survived dystentry, typhoid, pneumonia, TB, malaria etc
-
--
Today’s Europeans are the survivors of the 25%amongst which fibrogenicum is a super-survivor
F508deleted (recessive) CFTR carriage is common Randomly, every 50th - 80th European has it in single copy› this creates an unbiased source population sample › (genetic incidence of carriage is known in Europe)
BUT 2 copy carriage is lethal without therapy by 5yrs› Mortality-reducing Therapy is relatively affordable
› set the budget (see paper by Sims et al 2007 in the Lancet)
Phe508del/Phe508del
CZ
EuroCareCF partnered with European CF Registry ECFS
EU
Dundee
Data Centre
Legal
East data hub – Milan Macek
ECFSwww.eurocarecf.eu
Work Package 2
2006FP6
~30,000Registered
CF in Europe
Patients in 5-year Age Groups
12
1617
16
14
10
6
9
0
2
4
6
8
10
12
14
16
18
0-5y 5-10y 10-15y 15-20y 20-25y 25-30y 30-35y 35+y
Age Groups
Pe
rce
nta
ge
of
Pa
tie
nts
Each Bar shows a 5 year age band, except last one
0-5 yr 5-10 yr 10-15 yr 15-20 yr 20-25 yr 25-30 yr 30-35 yr >35 yr Total 3549 4676 4922 4795 3945 2768 1854 2584 29093
12% Pre-school
%
~1500 Missing pre-school
Firstschool
Teenagechildren
~3550
Patients in 5-year Age Groups
0
5
10
15
20
25
30
35
0-5y 5-10y 10-15y 15-20y 20-25y 25-30y 30-35y 35+y
Age Groups
Pe
rce
nta
ge
of
Pa
tie
nts
Patients in 5-year Age Groups
0
2
4
6
8
10
12
14
16
18
20
0-5y 5-10y 10-15y 15-20y 20-25y 25-30y 30-35y 35+y
Age Groups
Pe
rce
nta
ge
of
Pa
tie
nts
Patients in 5-year Age Groups
0
5
10
15
20
25
30
35
0-5y 5-10y 10-15y 15-20y 20-25y 25-30y 30-35y 35+y
Age Groups
Pe
rce
nta
ge
of
Pa
tie
nts
Patients in 5-year Age Groups
0
2
4
6
8
10
12
14
16
18
20
0-5y 5-10y 10-15y 15-20y 20-25y 25-30y 30-35y 35+y
Age Groups
Pe
rce
nta
ge
of P
atie
nts
Country A
Country B Country D
Country C
What Message does the pattern provide?
CF occurs randomly and yet expected pre-school number of about 5000 we only found 3500 Question: Where have the children gone?› Either they are undiagnosed› Or they have died
Let us look at when most CF patients are diagnosed
AIM: to find ‘remote’ answers of prognostic significance
Age at Diagnosis for all Patients
1112
8
6
11
7
9
6
43 2
54
2 21
0 0 0
9
0
2
4
6
8
10
12
14
0m 1m 2m 3m4-
6m7-
12m 1y 2y 3y 4y 5y6-
10y
10-1
5y
15-2
0y
20-3
0y
30-4
0y
40-5
0y50
+y
Not Rep
orted
Per
cen
tag
e o
f P
atie
nts
Let us expand the age at diagnosis….
Pre-schoolYet to be diagnosed >1500
Expanded age scale from birth
At Birth Diagnosis
Severe disease
All Europe patients
ALLCF
44%Common
Type
Phe508Del
homozygotes
96% by 10 yr98% by 15 years
2% Phe508delhomozygotes
Europe n=29093
Genotypes Reported - Percentages
11
29
60
F508del/F508del F508del/Other Other/Other
Genotypes Reported - Percentages
59
36
5
F508del/F508del F508del/Other Other/Other
Genotypes Reported - Percentages
51
33
17
F508del/F508del F508del/Other Other/Other
Country X Country ZCountry Y
F508del/F508del
F508del/Other
Other/Other
Minority Majority Even splitThis common form arose once, presents in childhood almost always creating a marker for CF diagnostic ability
and eliminates European geographic differences
Setting Up the Hypothesis on Demography
Nearly all should be diagnosed by 15 years› They have severe disease› They should have access to good care› They die if untreated
Very few will die if treated › …UK Data as an example
Males Females
Born in 1960-79
1989
Eur Respir J. 2007 29:522-6 …….Data from the UK100
How can the EU use this information at a country level to measure outcome?
Calculate the number with common CF (eliminate 98% of ascertainment bias)› Divide that Phe508del number by all CF
patients to get a percentage in childhood
Assume that <5% should die by 15 years of age (if care is good)
Look at the pattern of ages
Country 1 EU $£€
+++Country 2 NON-EU MONEY
~ +
See McCormick et al 2010 Lancet
57 54 54 5752 51
59
3940
2028 24
13
27
0 00
10
20304050
6070
0-5y
6-10
y
11-1
5y
16-2
0y
21-2
5y
26-3
0y
31-3
5y36
y+
Del
tF/A
ll
as p
erce
nt
% of severely affected patients by ageAll CF patients at same age
Country1 2
Severe-common CF to All-CF ratio by age
P<0.00001 P<0.000005
Patients in 5-year Age Groups
0
5
10
15
20
25
30
35
0-5y 5-10y 10-15y 15-20y 20-25y 25-30y 30-35y 35+y
Age Groups
Pe
rce
nta
ge
of
Pa
tie
nts
Patients in 5-year Age Groups
0
2
4
6
8
10
12
14
16
18
20
0-5y 5-10y 10-15y 15-20y 20-25y 25-30y 30-35y 35+y
Age Groups
Pe
rce
nta
ge
of
Pa
tie
nts
Patients in 5-year Age Groups
0
5
10
15
20
25
30
35
0-5y 5-10y 10-15y 15-20y 20-25y 25-30y 30-35y 35+y
Age Groups
Pe
rce
nta
ge
of
Pa
tie
nts
Patients in 5-year Age Groups
0
2
4
6
8
10
12
14
16
18
20
0-5y 5-10y 10-15y 15-20y 20-25y 25-30y 30-35y 35+y
Age Groups
Pe
rce
nta
ge
of P
atie
nts
Country A
Country B Country D
Country C
Age in years
0-<1
0
10-<
20
20-<
30
30-<
40
40-<
50
50-<
60
60-<
70
70-<
80
% c
ha
ng
e in
CF
po
pu
lati
on
s
ize
fro
m p
revi
ou
s d
ec
ad
e
-100
-90
-80
-70
-60
-50
-40
-30
-20
-10
0
10
20
30
Rising population
Falling population
1. McCormick et al Lancet 2010: sets the target2. Sims et al Lancet 2007: sets the costs3. Jagger et al Lancet 2008: sets the context
Age in years
0-<1
0
10-<
20
20-<
30
30-<
40
40-<
50
50-<
60
60-<
70
70-<
80
% c
ha
ng
e in
Fd
el5
08/F
de
l50
8
CF
po
pu
lati
on
siz
e f
rom
pre
vio
us
de
ca
de
-100
-90
-80
-70
-60
-50
-40
-30
-20
-10
0
10
20
30
All CFCommon Phe508deleted form
Western Europe,EU 2003
Non-EU
Now we have both a target and the costs
Rate of Change by succeeding decadeWe have a costed target
50,000 years ago0−2Κ
ΑD
∆F508
We are a CF legacy
2010
BCBC
-
10K10K
−10Κ
75% died before 6 years of age
Genetically fit
Fibrogenicum super-survived dystentry, typhoid, pneumonia, TB, malaria etc
-
--
We should thank patients with CF for suffering on our behalf
Treharne et al 2009 Febs Lett CF and smallpox
In Rare Diseases, Demographics are Powerful if…
Look at and use the age profile› Compare like with like apples with apples› i.e. common severe genes to measure the
severe/total ratios in childhood› i.e. divide severe by total and show this to be
stable (against first 15 years of life in CF)
Calculate the median age at death› Compare spread of medians with the most
robust registry data
Direct GeneticsX%X%
Environment
Z%Z%
Indirect genetics (mosaic of DNA inherited
from early man)Y%Y%
Inflamm bowelArthritis
SLE
Chromosome 6/Toll
Conclusion
Margaret Fraser…………………..………………….someone who knows data
Gita Mehta…………………………someone who knows project management
Milan Macek…………………someone who has a network (through genetics)
David Sheppard………………………………………….manage the resources
Funding……………………………………….………………EU FP6 EuroCareCF
www.eurocarecf.eu
www.cystic-fibrosis.org.uk
SKILL MIX IS THE KEY: J McCormick MD
RegisterSKILL MIX
Standards Legal Framework
EuroCareCFEuroCareCF Skill Mix – will differ
Single gene direct geneticsX%X%
Indirect genetics (mosaic of DNA inherited
from early man -still with us)Y%Y%
Environmental issues
Spectrum of disease presentationSpectrum of disease presentation
Z%Z%
None
Pre-natal death
Overview of Outcomes in rare diseases
1960-20002008- next 20 years
1970-2008
Single gene direct geneticsX%X%
Indirect genetics (mosaic of DNA inherited
from early man -still with us)Y%Y%
Environmental issues
Spectrum of disease presentationSpectrum of disease presentation
Z%Z%
None
Pre-natal death
Overview of Outcomes in rare diseases
1960-20002008- next 20 years
1970-2008