Transcript No Slide Title

Testing in the Rheumatic Diseases
Salahuddin Kazi, M.D.
Questions to Answer When Applying a Valid
Diagnostic Test to a Specific Patient*
• Is the test available, affordable, accurate and
precise in our setting?
• Can we generate a clinically sensible
estimate of our patient’s pre-test probability?
• Will the resulting post-test probability affect
our management and help our patient?
*Evidence Based Medicine: 2nd Edition, Sacket et al, 2000
Test Statistics - A Review
• Sensitivity - The proportion of affected
individuals with a positive test
• Specificity - The proportion of unaffected
individuals with a negative test
• Utility lies at the extremes - SpPin “High
specificity; positive test rules in” and SnNout
“High sensitivity; negative test rules out”
Negative
Test Result
Gold Standard
Positive
Negative
True Pos
False Pos
a
c
Positive Predictive
Value = a/ a+b
b
d
Negative Predictive
Value = d/ c+d
False Neg
True Neg
Prevalence = a+c/ a+b+c+d
Sensitivity
= a/ a+c
Specificity
= d/ b+d
Likelihood Ratios
• The likelihood that a given test result would
be expected in a patient with the target
disorder compared with the likelihood that
the same result would be expected in a
patient without the target disorder
• +LR = sensitivity/(1-specificity)
• -LR = (1-sensitivity)/specificity
Likelihood Ratio Normogram
0.1
99
0.2
0.5
1
2
5
10
20
30
40
50
60
70
80
90
95
1000
95
90
500
200
100
50
20
10
5
2
1
80
70
60
50
40
30
0.5
0.2
0.1
0.05
0.02
0.01
0.005
0.002
0.001
20
10
5
2
1
0.5
0.2
99
Pre-test
probability
0.1
Likelihood
ratio
Post-test
probability
Case #1
• A 32 y/o woman describes a 6 week history of pain
and stiffness in her hands
• No history of fever, rash, dysuria, conjunctivitis,
travel or exposure. No prior renal disease, seizures,
or serositis. Her mother has deforming arthritis.
• On exam there is warmth and soft-tissue swelling of
the 2nd and 3rd MCPs bilaterally, the 3rd right PIP,
and the left wrist. There seems to be a small
effusion in the left knee.
• Labs: SMA6 and U/A - normal; mild anemia; CRP
2.7; ANA - 1:40, diffuse; Rheumatoid Factor - 320
Rheumatoid Factor
• Anti-IgG - can be all Ig classes
• Specific for Fc portion of IgG
• Can be polyclonal (typical of autoimmunity)
or monoclonal (typical of lymphoid
malignancy)
• Causes immune complex damage
• Reported as “units” or titer
RF Test Characteristics
•
•
•
•
Sensitivity for RA is ~80%
Specificity is 85-95%
+LR of 5-16 depending on population studied
High titer is associated with more severe RA
with extra-articular manifestations
• Monitoring titer as an indicator of disease
activity is not appropriate
Conditions Associated with RF
• Normal individuals (5%), especially with age (15%)
• Rheumatoid arthritis (85%), Sjögren’s Syndrome,
SLE (25-50%)
• Viral Infections: Hepatitis C (25-50%),
mononucleosis, HIV, influenza
• Bacterial Infections: IE (25-50%), TB (10-25%),
leprosy, syphilis, brucellosis
• Parasites: Typanosomiasis, malaria,
schistosomiasis, etc.
• Other: Sarcoidosis, pulmonary fibrosis (10-25%),
chronic liver disease
Case #1 – Using the Test Results
• Chronic, inflammatory, symmetrical
polyarthritis of the hands in a young woman
• What’s your pre-test probability that this
patient has RA?
Highly positive RF takes a 50% pre-test probability to
a >95% post-test probability
0.1
99
0.2
0.5
1
2
5
10
20
30
40
50
60
70
80
90
95
1000
95
90
500
200
100
50
20
10
5
2
1
80
70
60
50
40
30
0.5
0.2
0.1
0.05
0.02
0.01
0.005
0.002
0.001
20
10
5
2
1
0.5
0.2
99
Pre-test
probability
0.1
Likelihood
ratio
Post-test
probability
In this case, a test
with a moderate
+LR makes the
diagnosis almost
certain in a patient
with a high pre-test
probability
Case #2
• A 64-year-old female was evaluated for generalized
joint pain and muscle pain, fatigue, fever and chills
for the past 6-8 weeks
• No rash, Raynaud’s, weight loss
• Graves’ disease 18 years ago - radioactive iodine
• Family history: SLE, thyroid disease
• PE: Tender joints but no joint swelling
• Labs:
– CBC, Chem 7, LFT’s, UA, TSH - all normal
– ESR 18 mm/h, CRP <0.8, RF negative, ANA
positive 1:80, homogeneous pattern
Anti-Nuclear Antibodies
• Began with the demonstration of the “LE
cell” by Hargraves in 1948
• Includes antibodies to a number of antigens,
including native DNA
• Performed by indirect immunofluorescence
• Reported as “negative” - usually less than a
certain titer, or as a titer and pattern
ANA - Characteristics
• Sensitivity - 95-100%
• Specificity - Depends on titer used as cut-off
– 15-30% of normals have ANA of 1:40
– 5% have ANA of 1:160
• +LR is ~20; utility for SLE is based on prevalence:
– General population 50/100,000
– Young, African-American women 400/100,000
– Children/elderly men 1/100,000
Immunofluorescence Patterns of ANAs
Pattern
Homogeneous
Peripheral or Rim
Speckled
Nucleolar
Centromere
Cytoplasmic
Related Antigen Specificities
Chromatin, Histone, DNA, Ku
DNA, Lamins
RNP, Sm, Ro, La, Ku
Topoisomerase I (Scl-70)
RNA Pol 1, Fibrillarin, PM-Scl
CENPs
Ribosomal P, Aminoacyl t-RNA synthetases
Homogeneous
Rim or Peripheral
Nucleolar
Speckled
Causes of a Positive ANA
Rheumatic Disease
SLE
Percentage
Positive
>95%
Disease Specific Ab’s
Systemic Sclerosis
60%-90%
Sjögren’s Syndrome
75%
Anti-Sm,
Anti-dsDNA
Anti-centromere
Anti-Scl-70
Anti-Ro, Anti-La
Mixed CTD
95%-99%
Anti-RNP
Poly/Dermatomyositis
25%
Anti-Jo-1
RA
15%-35%
Rheumatoid Factor
Interpreting a Positive ANA
Pretest likelihood of
lupus
ANA titer
Action
Low
 80
Ignore
 160
Observe; look for an
alternative explanation
Observe; look for an
alternative explanation
Check for diseasespecific antibodies
Observe; look for an
alternative explanation
Check for diseasespecific antibodies
Moderate
 80
 160
High
Negative
Positive, any
titer
Other Causes of Positive ANA
Non Rheumatic Disease
Infections
Inflammatory bowel
disease
Autoimmune hepatitis
Pulmonary fibrosis
Endocrine diseases
Hematologic diseases
Neoplastic diseases
End-stage renal disease
Post-transplant
Healthy People
Pregnancy
Older people
Family history of
rheumatic disease
Drug induced
Case #2:Why is the ANA Positive?
•
•
•
•
•
History and PE: Does not suggest a CTD
Labs: normal except for positive ANA
Pretest probability of SLE is low
Posttest probability for SLE remains low
Look for an alternative explanation
– Elderly female
– Positive family history of rheumatic disease
• Reassure: ANA result is a normal finding
0.1
99
0.2
0.5
1
2
5
10
20
30
40
50
60
70
80
90
95
1000
95
90
500
200
100
50
20
10
5
2
1
80
70
60
50
40
30
0.5
0.2
0.1
0.05
0.02
0.01
0.005
0.002
0.001
20
10
5
2
1
0.5
0.2
99
Pre-test
probability
0.1
Likelihood
ratio
Post-test
probability
Although an ANA
>1:160 has a high
+LR, it should not
be used to screen
patients without
clinical evidence of
autoimmune
disease
Ordering an ANA
• To confirm the diagnosis of SLE when the
clinical suspicion is high
• To exclude SLE when the clinical suspicion
is moderate (2 or 3 lupus criteria)
• Avoid ordering it when the clinic suspicion
for SLE is low - a positive result can cause
diagnostic confusion and unnecessary
anxiety
Anti-DNA Antibodies
• Detect antibodies to native (double
stranded) DNA
• Typical methods are ELISA and
immunofluorescence on Crithidia
• Can have both diagnostic and
prognostic significance
Anti-DNA - Characteristics
Sensitivity - 60% for SLE
Specificity - 97%
Low titers seen in 2-5% of RA, Sjögren’s,
scleroderma, relatives of SLE pts., etc.
Average +LR of 16 and -LR of 0.49 means that a
positive anti-DNA has a large impact, but lack
of one doesn’t exclude SLE
Anti-DNA - Prognosis
• SLE Disease activity: Useful, but with small
+LR (~4)
• Nephritis: Associated, but with very small
+LR (~1.7)
• Rising titers may predict a flare of disease
activity in some, but not all, patients
• Clinical correlation is advised
Anti-ENA
• Small nuclear RNP
– Sm: Seen in 15-30% of SLE; specific
– U1-RNP: 30-40% of SLE; also RA, Sjögren's,
scleroderma, and overlap syndromes
• Anti-Ro and anti-La
– Subacute cutaneous LE
– Sjögren's syndrome
– Neonatal lupus with congenital heart block
“ENA”-Extractable Nuclear Antigens
RNP Antigens
Sm Antigens
C
70kDa
28kDa (B) 28kDa (B’)
G
16kDa (D)
33kDa (A)
3’
5’
U1RNA
E
F
Anti-Scl-70/Anti-Centromere
• Scl-70 = Topoisomerase I; seen in 40-70% of
patients with diffuse scleroderma; worse
prognosis with more organ involvement
• Centromere - 70-85% of patients with limited
scleroderma; associated with Raynaud’s
syndrome
• Neither is diagnostic by themselves
Case #3
• A 48-year-old male has chronic sinusitis with
occasional bloody drainage
• You order a c-ANCA
– Positive at 1:80
• The chest radiograph, creatinine and
urinalysis are normal
• What is the likelihood that he has Wegener’s
granulomatosis?
Positive Predictive Value of ANCA
4
Positive Predictive Value
100
1
2
3
1. Documented WG
2. Pulmonary-Renal Syndrome
3. Systemic Necrotizing Vasculitis
4. Rapidly Progressive GN
5. GN
6. Hospitalized Patient
5
50
This Patient
0
6
50
100
Disease Prevalence
Jeanette: Amer J Kidney Dis 18:164, 1991
0.1
99
0.2
0.5
1
2
5
10
20
30
40
50
60
70
80
90
95
1000
95
90
500
200
100
50
20
10
5
2
1
80
70
60
50
40
30
0.5
0.2
0.1
0.05
0.02
0.01
0.005
0.002
0.001
20
10
5
2
1
0.5
0.2
99
Pre-test
probability
0.1
Likelihood
ratio
Post-test
probability
Wegener's is rare
(~0.4/100,000).
Without signs of
progressive,
necrotizing vasculitis,
even a test with a high
likelihood ratio is not
helpful
ANCA Characteristics
• C-ANCA (Proteinase-3)
– 90% specificity and 50-90% sensitivity for
active Wegener's granulomatosis
• P-ANCA
– MPO - 60% of microscopic polyangiitis, ChurgStrauss
– Cathepsins, lactoferrin, elastase
• Should not take the place of tissue biopsy
ANCAs and Rheumatic Autoimmune Diseases
• P-ANCA (not directed against MPO) reported in:
– RA, SLE, PM/DM, Sjögren's syndrome, Juvenile
chronic arthritis , Reactive arthritis, Relapsing
polychondritis*
• C-ANCA
– very rare in these diseases
• ANCA is not associated with increased
frequency of vasculitis in the autoimmune
rheumatic diseases
*Ann Intern Med 126:866-873, 1997
ANCA and Inflammatory Bowel Disease
• P-ANCA and some atypical patterns (not
directed at MPO)
– Ulcerative colitis - 40% to 80%
– Crohn’s Disease - 10% to 40%
• Does not facilitate the differential diagnosis
of patients with inflammatory bowel disease
• Correlation of titers with disease activity is
not sufficiently reliable
How are ANCAs detected?
• Indirect immunofluorescence (IIF)
– c-ANCA or p-ANCA pattern
• Enzyme linked immunosorbent assay (ELISA)
– specific antigens detected
– PR3: (c-ANCA on IIF)
– MPO: (p-ANCA on IIF)
• ANCA testing is problematic because of lack
of standardization between laboratories
Summary
• Connective tissue diseases have a low
prevalence
• Unselected “screening” of patients with
“arthritis panels” will result in large numbers
of false positives
• Estimation of clinical pre-test probability and
the knowledge of test characteristics are
useful tools for rationally ordering and
interpreting the results of diagnostic tests