Radiographic scoring in rheumatoid arthritis - The basics

1
Radiographic scoring in rheumatoid arthritis -
The basics

• Tuulikki Sokka, MD, PhD
• tuulikki.sokka_at_ksshp.fi

2
Learning Objectives

• History of different scoring methods
• Basics of the most often used methods
• Interpretation of radiographic scores in clinical
  trials
• Clinical use of radiographs
• Radiographic outcomes in selected clinical cohorts

3
History the main methods

• Steinbrocker 1949
• Kellgren 1956
• Sharp 1971
• Van der Heijde modification
• Larsen 1973
• modifications

4
Steinbrocker method

• Stage I - IV
• Relates to anatomic stages
• radiographs of handswrists
• The grade is determined by the worst change in
  any joint
• Limitations narrow scale bias toward the most
  severely affected joint

5
Kellgren method

• 0-4, based on standard set of radiographs
• global one grade is given as a summation of
  abnormalities for all the joints in both hands
  and wrists
• Limitations narrow scale weighted to reflect
  the most damaged joints

6
Sharp method (1)

• Purpose to develop a quantitative assessment for
  radiographic changes in RA
• Included Handswrists

7
Sharp method (2)

• Initially, 10 features were analyzed
• Periosteal reaction
• Cortical thinning
• Osteoporosis
• Sclerosis
• Osteophyte formation
• Defects
• Cystic changes
• Surface erosions
• Joint space narrowing
• Ankylosis

• Reason to delete items
• Rare
• Technical problems
• Secondary changes

8
Sharp method (2)

• Initially, 10 features were analyzed
• Periosteal reaction
• Cortical thinning
• Osteoporosis
• Sclerosis
• Osteophyte formation
• Defects
• Cystic changes
• Surface erosions
• Joint space narrowing
• Ankylosis

Rare Technical problems Secondary changes
INCLUDED Erosion score Joint space narrowing
9
Sharp method (3)

• Erosion score principles
• Score 0-5 for each joint
• one point for each erosion in each joint and 5
  for total destruction
• 29 areas were analyzed in both handswrists
  maximum possible score 290

10
Sharp method (4)

• Joint space narrowing score principles
• 0 - normal
• 1 - focal narrowing
• 2 reduction of lt50 of joint space
• 3 reduction of gt50 of joint space
• 4 ankylosis
• 27 areas in hands and wrists max score 216

11
Sharp method (5)

• How many joints? (1985)
• Factors to be considered
• Frequency of involvement
• Technical factors
• Minimum number of joints required in a patient
  population from mild to severe disease
• 17 for erosions
• 18 for joint space narrowing
• .. Still to decrease.

12
Van der Heijde modification of the Sharp score

• PRINCIPLES
• Feet included
• Number of hand joints decreased
• Scoring for erosions defined

13
The Sharp/van der Heijde Joints to be scored for
erosions
14
The Sharp/van der Heijde Joints to be scored for
joints space narrowing
15
Sharp van der Heijde method (1)Erosions

• Scoring of the hands 16 areas included
• Score 0-5 per joint
• 1 for discrete erosions
• 2-3 for larger erosions depending of the surface
  area involved
• 4 if erosion extends over middle of the bone
• 5 for complete collapse

16
Sharp van der Heijde method (2)Erosions

• Scoring of the feet 10 MTP and 2 IP joints of
  big toes
• Score 0-5 per each side of the joint total 0-10
• 1 for discrete erosions
• 2-3 for larger erosions depending of the surface
  area involved
• 4 if erosion extends over middle of the bone
• 5 for complete collapse

17
Sharp van der Heijde method (3)JSN, hands, feet

• Joint space narrowing score 15 areas for hands,
  6 for feet
• 0 - normal
• 1 - focal narrowing
• 2 reduction of lt50 of joint space
• 3 reduction of gt50 of joint space
• 4 ankylosis

18
Sharp van der Heijde method (4)

• Total scores
• Erosion scores for hands 160
• Erosion scores for feet 120
• JSN for hands 120
• JSN for feet 48
• Total 448

19
Larsen score (1)

• Background was a clinical observation
• A man with RA Steinbrocker 4 running to a bus
• Steinbrocker 4 is maximal damage
• Max damage and running to a bus do not match
• A better scoring method needed

20
Larsen score (2)

• Reference films for each joint
• Score 0-5 for each joint
• Scoring includes JSN and erosions
• Articular osteoporosis and soft tissue swelling
  were initially included but omitted later

21
Larsen score (3)

• Which joints?
• Scott 1995 10 PIPs, 10 MCPs, 10 MTPs, wrists
  multiplied by 5 total score 200
• Kaarela Kautiainen 1997 10 MCPs, II-V MTPs,
  wrists not multiplied total score 100

22
Larsen 0-100
23
Larsen scoring
24
Larsen vs. Sharp

• Are significantly correlated
• Pincus et al. J Rheumatol 1997
• Larsen less time-consuming and easier
• overall scoring for each joint
• wrist analyzed as one joint
• lower number of joints

25
Smallest Detectable Difference SDD

• SDD is the smallest change that can be reliably
  discriminated from the measurement error of the
  scoring method
• SDD is based on defining measurement error and
  95 limits of agreement
• Sharp vd Heijde on scale 448 SDD 5
• Larsen on scale 200 SDD 5.8

Bruynesteyn et al. AR 2002
26
Minimal Clinically Important Difference MCID

• MCID progression with the highest combined
  sensitivity and specificity for detecting
  relevant progression
• Sharp vd Heijde on scale 448 MCID 4.6
• Larsen on scale 200 MCID 2.3
• In both, roughly 1 of the maximum

Bruynesteyn et al. AR 2002
27
Radiographic scores in RCTs -interpretations
28

Radiographic progression in selected clinical
trials
Yazici Y, Yazici H, Arthritis Rheum 200654(supl)
29
Low radiographic damage in current RCTs Table
3. Change from baseline in disease
characteristics in the ITT population after 2
years of treatment in the TEMPO trial
MTX (n 206) Etan (n 202) Etan MTX (n
212) Year 2    Total Sharp score
(0-448)        Mean (95 CI) 3.34 (1.18,
5.50) 1.10 (0.13, 2.07) -0.56 (-1.05,-0.06)     
  Median (IQR) 0.00 (-0.11, 2.33) 0.00 (-0.66,
1.08) 0.00 (-1.41, 0.05)
vdHeijde AR2006
30
Few patients have radiographic damage in current
RCTs Total Sharp vdHeijde score (0-448) in the
TEMPO trial over 2 years
vdHeijde AR2006
31
Measures of RA over time short term vs. long term

• Long term
• Years - decades
• Deformities
• Radiographic damage
• Joint replacements
• Functional capacity
• Comorbidity
• Work disability
• Costs
• Mortality
• measures of outcomes

• Short term
• Months - years
• Swollen joint count
• Tender joint count
• ESR, CRP
• Pain
• Functional capacity
• Global health by patient
• Global health by Dr
• (Radiographic damage gt1yr)
• measures of disease activity

Clinical cohorts, longitudinal observational
studies, databases
RCTs
32
Radiographs clinical use
33
Two clusters of measures in RA
x-rays HAQ
joint deformity disease duration
pain RF joint tenderness
joint swelling ESR, CRP age HLA-DR4 pa
tient global work disability mortality

Pincus, Sokka. Best Pract Res Clin Rheumatol.
2003
34
The HAQ, CLINHAQ, or MDHAQ Patient Questionnaire
is Best Predictor in RA of

• Functional status (Pincus et al. Arthritis Rheum.
  1984, Wolfe et al. J Rheumatol. 1991)
• Work disability (Borg et al. J Rheumatol 1991,
  Callahan et al. J Clin Epidemiol. 1992, Wolfe and
  Hawley. J Rheumatol. 1998, Fex et al. J Rheumatol
  1998, Sokka et al. J Rheumatol 1999, Barrett et
  al. Rheumatology 2000, )
• Costs (Lubeck et al. Arthritis Rheum. 1986)
• Joint replacement surgery (Wolfe and Zwillich.
  Arthritis Rheum. 1998)
• Death (Pincus et al. Arthritis Rheum. 1984, Ann
  Intern Med.1994, Wolfe et al. J Rheumatol 1988,
  LeighFries J Rheumatol 1991, Wolfe et al.
  Arthritis Rheum. 1994, Callahan et al. Arthrits
  Care Res 1996, 1997, Soderlin et al. J Rheumatol
  1998, Maiden et al. Ann Rheum Dis 1999, Sokka et
  al. Ann Rheum Dis 2004)

35
Larsen Thoen Scand J Rheumatol 1987
100 75 50 25 0
Damage score 0-100
0 2 4 6 8 10 12 14
16 18 20 22 24
Disease duration, years
36
Fuchs et al. J Rheumatol 1989
100 75 50 25 0
Erosion score 0 - 4.33
0 2 4 6 8 10 12 14
16 18 20 22 24
Disease duration, years
37
Salaffi Ferraccioli Scand J Rheumatol 1989
100 75 50 25 0
Erosion score 0 - 150
0 2 4 6 8 10 12 14 16
18 20 22 24
Disease duration, years
38
The Jyväskylä Experience

• The Central Finland RA register includes all
  patients with diagnosis of RA since 1980s
  prospective in all patients since 1996
• 2,900 patients 2,300 alive
• Covers a population of 265,000

39
The North Pole
40
Jyväskylä Central Hospital is the only
rheumatology clinic in the Central Finland
District and serves a population of 265,000 2
full-time rheumatologists and 1 trainee 4 other
rheumatologists
41
The Central Finland RA Register

• Patient demographics
• History of onset of RA
• Classification criteria
• Extra-articular features
• Comorbidities
• Relevant surgeries
• All previous and present DMARDs

42
Patients with early arthritis

• All new patients with RA are included about 100
  early RA patients each year
• Baseline data includes patient self-report
  questionnaires, structured clinical status,
  laboratory tests, radiographs of hands and feet

43
Patient Monitoring in early RA since 1997

• Regular out-patient visits in rheumatology unit
  for 2 years
• A control visit at 1, 2, 5, and 10 years
  including patient self-reported outcomes,
  structured clinical status, update of RA register
  information, laboratory tests including RF and
  aCCP, and radiographs of hands and feet

44
Patient Monitoring

• Each visit, every patient is asked to complete an
  extended 2-page HAQ or self-report on a touch
  screen / GoTreatIT
• Rheumatologist a status form / GoTreatIT
• An annual mailed questionnaire to all patients in
  the RA Register since 1998
• A 5-year follow-up of 2000 population controls in
  2000-2005 2007

45
Radiographic outcomes in selected clinical cohorts
46
Radiographic outcomes over 5 years in 3 Jyvaskyla
Cohorts

• Patients with early RA
• 1983-85
• 1988-89
• 1995-96

47
1988-89
1983-85
1995-96
Larsen scores of RF patients over 5 years
Each line illustrates Larsen score of each patient
Sokka et al. J Rheumatol 2004
48
1983-85 DMARDs over 5 years Increasing use over
time
1988-89
Sokka et al. J Rheumatol 2004
1995-96
49
Radiographic outcomes of RF patients over 5
Years in 3 cohorts of patients with early
RA. 1983-85 1988-89 1995-96 N 46 53 3
8 Patients with an erosive disease at 5 years,
86 67 73 Patients with Larsen
gt10, Baseline 9 0 3 2 years 40 20 8
5 years 55 33 14
Patients in the most recent cohort have potential
for an erosive disease but the extent of damage
remained low compared to earlier
cohorts. Sokka et al. JRheumatol 2004
50
Radiographic outcomes in two cohorts

• The Heinola Cohort 103 patients with early RA in
  the 1970s
• The Jyvaskyla Cohort 85 patients with early RA
  in the 1980s
• All RF
• 8-year follow-up

51
Larsen score in the Heinola Cohort vs. Jyvaskyla
Cohort over 8 years
26
12
Disease duration (years)
Sokka T, Kaarela K, Mottonen T, Hannonen P. Clin
Exp Rheumatol 1999
52
Increased use of DMARDs in the later
cohort Heinola 1973-75 103 patients Early RA RF
Jyvaskyla 1983-89 85 patients Early RA RF saw
tooth strategy Sokka et al CER 1999
53
Median values with 95 confidence intervals for
the Larsen score in patients with lt5, 5-15 and gt
15 years of disease in 1985 and 2000 in TPclinic
Pincus, Sokka, Kautiainen AR 2005
54
Contemporary DMARDs in the 1985 Cohort
Pincus, Sokka, Kautiainen AR 2005
55
Contemporary DMARDs in the 2000 Cohort
Pincus, Sokka, Kautiainen AR 2005
56
Scoring of x-rays in RCTs vs. in clinical care

• Experienced assessors read x-rays
• Observers blinded to clinical data
• Observers blinded to the order of radiographs
• Strict methodology to get accurate scores

• Every clinician to have basic knowledge about
  x-rays
• X-rays add to clinical data
• Serial x-rays to be compared to detect
  progression/improvement
• Understanding of radiographic progression

57
32nd Scandinavian Congress of Rheumatology 30
January - 3 February 2008 Levi, Lapland, Finland
Further information www.congrex.fi/scr2008
58
(No Transcript)
59
To read

• van der Heijde D. How to read radiographs
  according to the Sharp/van der Heijde method. J
  Rheumatol 1999 26743-745.
• Kaarela K, Kautiainen H. Continuous progression
  of radiological destruction in seropositive
  rheumatoid arthritis. J Rheumatol 1997
  241285-1287.