Transcript Document

Thought barriers
to understanding
rheumatic diseases:
Halsted Holman revisited
Hasan Yazıcı
University of Istanbul
Disclosers: none
Plan
1.
What did Dr. Holman say?
2.
•
•
•
Further thought barriers:
“lumping” of diseases
lack of proper controls in data collection
the popularity of inductive reasoning, leading to:
- the urge to prove rather than to falsify
3.
An audit of global Behçet syndrome research in the last decade
Thought Barriers
I.
Cross sectional observations in chronic
diseases & emphasis on acute disease
II. Reductionism
(the single cause or aberration)
III. Abnormalities are always harmful.
HR Holman Arthritis Rheum, 1994
Thought Barriers
I.
Cross sectional observations in chronic
diseases & emphasis on acute disease
II. Reductionism
(the single cause or aberration)
III. Abnormalities are always harmful.
HR Holman Arthritis Rheum, 1994
Breast cancer in Riley’s mice
• Only C3H mice who are nursed by C3H mice get the
tumor at 1 year of age. (genetic)
• It only develops if the mice are infected with the
Bittner virus during suckling. (environmental)
• If the suckling mice live in standard conditions the
median time to tumor development is 364 days.
• If the suckling mice live in improved conditions the
median time to tumor development is 565 days.
V Riley, Science 1975
Thought Barriers
I.
Cross sectional observations in chronic
diseases & emphasis on acute disease
II. Reductionism
(the single cause or aberration)
III. Abnormalities are always harmful.
HR Holman Arthritis Rheum, 1994
HR Holman Arthritis Rheum, 1994
Further Issues
•
•
•
•
What maintains chronicity?
Target organ variability
Inter and intra patient disease severity
Discordance between clinical and laboratory
findings
• Variation in age of onset in “genetic” disease
• Variation in response to therapy
• How does the disease go away?
Additional thought barriers
• Lumping
• Lack of attention to the proper control
groups
• Inductive reasoning leading to the urge
to prove, not to falsify
Some “lumps”
• Autoimmune diseases
• Seronegative
spondarthritides (AS vs
RA)
• Th1 and Th 2 cytokine
response
• Paucimmune vasculitis
• Autoinflammatory
diseases
Some “lumps”
• Autoimmune diseases
• Seronegative
spondarthritides (AS vs
RA)
• Th1 and Th 2 cytokine
response
• Paucimmune vasculitis
• Autoinflammatory
diseases
HLA-B27 in RA
HLA- B27+
• among 6/32 (19%) patients with RA
• Among 3/50 (6%) healthy controls
H Yazici et al. Ann Rheum Dis, 1987
HLA B27 and RA
• RA is increased among the relatives of
patients with AS - B Haar et al. Scand J
Rheumatol 1987
• HLA – B27+: 25% among female patients
with RA and 5% among the controls -
MA Sheritt et al. Hum Immunol 1996
• HLA- B27 and posterior uveitis
-A Rodriguez et al. Ophthalmology 1994
-A Anshu et al. Ophthalmologica 2007
SIR’s when a parent has RA
Siblings
RA
3.45
AS
3.03
Scleroderma (loc.) 2.93
Sjögren’s
2.54
SLE
2.36
Hashimoto/hypoth 1.78
y.
Pernicious anemia 1.74
PR
1.41
Asthma
1.37
Offspring
3.02
2.96
2.40
2.25
1.65
1.54
1.53
1.32
1.32
Hemminki K et al. Arthritis Rheum 2009
Target organ associations in BS
among of 272 consecutive patients:
• Factor I :
- oral ulcers + genital ulcers+ erythema nodosum
• Factor II:
- superficial vein thrombosis + deep vein thrombosis
• Factor III:
- uveitis
• Factor IV:
- acne + arthritis
R Tunc et al. 2002 J Rheumatol
Enthesitis in BS
G Hatemi et al. Arthritis Rheum, in print
SI disease and HLA B27 in BS
with Acne & Arthritis
I.
SI disease + (blind assesment with 3
observers)
•
•
•
•
AS: 16/16
BS with AA: 6/30
BS without: 2/27
RA: 3/12
II.
HLA B27 +
•
•
BS with AA: 2/27
BS without: 0/17
G Hatemi et al. Arthritis Rheum, in print
21
Fact or Myth
• Behçet’s syndrome is an
autoinflammatory condition.
Issues
•
•
•
•
•
BS
BS
BS
BS
BS
is rare in childhood.
does not run with febrile episodes.
does not run with polyserositis.
does not run with skin rashes.
does not respond well to colchicine.
Web resources for rare autoinflammatory diseases: towards a
common patient registry
I Touitou et al. Rheumatology , 2009
Item frequency in six auto-inflammatory patient registries
Touitou, I. et al. Rheumatology 2009 48:665-669;
doi:10.1093/rheumatology/kep056
Copyright restrictions may apply.
Percentage of items present in each registry among the 130 recorded
Touitou, I. et al. Rheumatology 2009 48:665-669;
doi:10.1093/rheumatology/kep056
Copyright restrictions may apply.
Additional thought barriers
• Lumping
• Lack of attention to the proper control
groups
• Inductive reasoning leading to the urge
to prove, not to falsify
What is FMF?
• “FMF is an automasal recessive disease
caused by mutations in the MEFV
gene.”
J Irrep Res 1998-
FMF - MEFV
• The phenotype is not homogenous
• 1/4 do not carry any MEFV alleles
• There are FMF families in which there is
dominant transmission
• MEFV polymorphisms are observed in
many inflammatory diseases
Issues
• About 20% of the patients do not carry any
mutations of MEFV.
• About 1/3 of the patients bear a single mutation.
• FMF families with a true, dominant inheritence is
well documented.
• There are patients who present with various
clinical manifestations not typical of FMF and
who carry MEFV mutations either as
heterozygote, compound heterozygote or as
complex alleles.
Control Groups in Genetic
Association Studies
Articles
(n=264)
Studies with only healthy
control groups
211 (79.9%)
Studies with healthy and
diseased control groups
24
Studies with only diseased
control groups
Family-based studies
(9.1%)
2 (0.76%)
27 (10.2%)
F Esen et al. Clin Exp Rheumatol, 2009
Referee comments
• So when they suggest: 'Had the initial FMF-pyrin work included
diseased controls, for example patients with other auto or
otherwise inflammatory conditions, the students of FMF would
have been where they are now quite a number of years ago.' I
would respond that had the two FMF consortia used a Behçet or
Crohn's disease control group, they might still search for the
gene now…”
Additional thought barriers
• Lumping
• Lack of attention to the proper control
groups
• Inductive reasoning leading to the urge
to prove, not to falsify
Development of the BD-QoL: a quality of life
measure specific to Behçet’s disease. G
Gilworth et al. J Rheumatol, 2004
• 71 QoL items are tabulated by qualitative in-depth
interviews with a limited number of patients.
• 30 non-redundant items are selected by the Rasch
model.
• The selected items are validated in a larger group
of BS patients.
Items best fitting the Rasch
model
•
•
•
•
•
It is difficult to get out of bed.
My life revolves around hospital visits.
My condition is drastically affecting my life.
I often get frustrated.
My condition affects important decisions in my
life.
G Gilworth et al. J Rheumatol, 2004
Issues
• No specificity data
• Responsive to change?
• The actual questions are not available.
The aim to prove
Pyrin levels with 1 versus 2
mutations
Pyrin levels as compared to
diseased controls
An Audit of Behçet’s Syndrome
Research: A 10-year Survey
F. Esen1, E. K. Schimmel2, H. Yazıcı1, Y. Yazıcı3
1.
İstanbul University, Cerrahpaşa Medical School, Department of Medicine, Division of
Rheumatology
2.
University of Missisipi Medical Center
3.
New York University School of Medicine, NYU Hospital for Joint Diseases42
Methods (1)
• PubMed Search: (Behcet) and (“Journal
Name”[Jour])
Limits: published in the last 10 years,
English, Date: 21.01.2009
• Search was repeated 60 times for articles
in the 15 highest impact factor
Rheumatology, Ophthalmology,
Dermatology and Medicine Journals.
Methods (2)
• 602 articles were found.
• Only the original research articles are
included.
• Case reports, letters, reviews and the
studies where the aim of the study was
not researching BS were excluded.
• 280 articles qualified for the survey.
44
Specialty Journals
45
Countries
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
46
Turkey
Japan
Korea
Italy
UK
Israel
Germany
France
China
Tunisia
USA
Greece
Spain
Netherlands
Iran
Saudi Arabia
Switzerland
Bulgaria
Chile
Egypt
Lebanon
Morocco
Singapore
Taiwan
Article #
107
30
25
19
17
16
12
9
7
7
6
5
4
3
2
2
2
1
1
1
1
1
1
1
Study Design (Time Element)
47
Longitidunal Studies
• Clinical descriptive: (1/45)
• Clinical (prospective, hypothesis testing): 1/16
• Mainly laboratory (human): 4/91
Y Yazici et al. ongoing
Study Design (Control Groups)
49
Diseased Control Groups
50
Power Calculations
51
Power Calculations
• Clinical descriptive: (0/45)
• Clinical (prospective, hypothesis
testing): 0/16
• Mainly laboratory (human): 0/91
Conclusions (BS audit)
• Most of the studies (83%) are cross-sectional,
although the disease is chronic.
• Prospective longitudinal studies are few (7%).
• The use of diseased control groups is not
(surprisingly (!) rare (42%), but still needs to be
improved.
• Basic demographic data is missing in a considerable
number of papers (21%).
• Power calculations are very rare (3%), even in RCTs
(17%) and are not even considered in clinical
hypothesis testing.
53
Summary
• As Dr. Holman aptly pointed out years
ago, there are distinct subheadings
under unfruitful thinking in medicine.
• The urge “to prove” might be the
common denominator in such thinking.
“Meanwhile much of our society at large is
busy revisiting the days of the Inquisition v
Galileo. The wars of religion are back,
superstition threatens our schools, and
Bible-thumpers preach that Darwin got it
wrong. Our modern heritage of reason,
formed in the Enlightenment, is becoming
eclipsed by what a cynic might call the
Endarkenment.”
G. Weismann in Galileo’s Gout, 2007