General introduction: Consecutive pathological phases have been recognized in the course of interphalangeal (IP) finger joint OA: a non-erosive OA joint (“S” or stationary OA joint) can enter the “J” phase when the joint space disappears, and then the “E” phase when manifest erosive changes occur. Next, the affected IP joints show signs of repair or remodelling and the “R” phase ensues (6).
On the radiographs, erosive changes in the affected IP joints include the disappearance of parts of the joint space followed by or concurrently with the appearance of substantial osteolytic areas in the subchondral bone and the subchondral plate. Alternatively, a subchondral plate collapse without complete disappearance of joint space may occur. In the end, the joint space of the affected IP joint will appear destructed and enlarged. Destructive phases, however, are always followed by repair or remodelling. Then, new irregular sclerotic subchondral plates are formed and a new joint space becomes visible, subchondral osteolytic areas gradually disappear and huge osteophytes are formed. No further evolution is seen in remodelled IP joints.
A five plate atlas containing 50 radiographs illustrate the pathological changes in the subchondral bone plate and synovial joint space (40 images), and the destruction of the subchondral bone area (10 images) of proximal interphalangeal (PIP) finger joints during their progression through the consecutive pathological phases. Changes in the structure of the subchondral plate and the synovial space, and of the subchondral bone area were indicated by illustrative line drawings and commented. Analogous pathological changes occur in distal IP finger joints.
Subchondral bone plate and synovial joint space. Plates 1 to 4 contain 40 images that illustrate the pathological changes in the subchondral bone plate and the synovial joint space in the PIP finger joints. When the joint space is still identifiable, the subchondral bone plates on these radiographs appear as well defined regular linear margins flanking the synovial joint space. Usually the subchondral bone plates are seen as brighter radioopaque linear structures. Subchondral bone plates covering the top of the proximal phalangeal bone, however, may not necessarily appear as a brighter radioopaque linear structure (image 1-10 on plate 1). The joint space is recognized as a radiotranslucent area bordered with 2 subchondral plates.
The top rows show the proximal (and distal) IP joints of a single digit with faint dotted vertical lines indicating the synovial joint space width.
On the middle rows line drawings locate the remaining subchondral plate together with 2 reference lines - confined by the vertical dotted lines shown in the top row - indicating the total length of 2 subchondral plates visible on these two dimensional images. Numbers indicate the percentage amounts of subchondral bone plate detected: length of the line drawings in proportion to the length of the reference lines.
On the bottom rows, line drawings locate the remaining synovial joint space according to the aforementioned definition. One single reference line indicates the theoretical width of the synovial joint space. Numbers indicate the amount of subchondral joint space (length of the line drawings) detected in proportion to the reference length.
Plate 1 displays 1 normal PIP joint (image 1), 3 non-erosive osteoarthritic (OA) PIP joints (image 2-4), and 6 erosive PIP finger joints in the “J” phase of which the synovial joint space in part had disappeared (image 5-10). Line drawings locate the remaining subchondral bone plate and synovial joint space. Normal joints and non-erosive IP finger joints show uninterrupted subchondral bone plates not in contact with each other with in between a continuous synovial space: subchondral bone plates and synovial joint space are 100% intact. Image 5 to 10 show IP joints with decreasing proportions of normal subchondral plate and synovial joint space. In these IP joints in the “J” phase the subchondral bone plates partially contact each other and a fraction of the plate is not recognized as such. Part of the synovial joint space had disappeared. In case the joint space has completely disappeared (image 10) the subchondral bone plates entirely merge. One single radioopaque line is then seen across the joint and the amount of subchondral bone plate detected with this system is 50%.
Plate 2 displays 10 PIP finger joints in the “E” phase with subchondral plate structures and synovial joint spaces that had disappeared and pronounced osteolytic activity in the subchondral bone area. The PIP finger joints are ordered with increasing proportions of structured subchondral bone plate. However, no synovial joint space is identified in these joints but for the joint on image 20 where 2 subchondral bone plates flanking a radiotranslucent zone over a short distance are observed.
Images 11 and 12 show PIP finger joints where the entire subchondral bone plate has collapsed leaving an eroded area in place. Image 13 and 18 show PIP finger joints with only just 10 and 40% of subchondral bone plate is left on the distal end of the proximal phalangeal bone. The opposed subchondral plate has been destroyed. According to the abovementioned definitions, no synovial space is observed in these joints as flanking subchondral bone plates have disappeared. Images 14 to 19 show PIP joints where the joint space has disappeared and parts of the remaining subchondral bone plates show up as one single brighter radioopaque structure. Proportions of subchondral plate increasing from 10 to 50% are seen in the PIP joints in image 13 to 19. Structures are more difficult to distinguish in these “E” phase IP finger joints and inter- and intrareader reproducibilities will be the lowest when assessed in these pictures.
Plate 3 displays 10 PIP finger joints, most of them in the “E” phase, except 25 and 27 being classified as “J” phase joints, and 28 and 29 to be classified as “R” joints. The PIP finger joints are ordered with increasing extent of subchondral bone plate, ranging from 50 to 90%. The amount of synovial joint space observed increases with the increasing proportions of subchondral plate.
Plate 4 displays 10 PIP finger joints. Image 31 shows a PIP finger joint progressing from the “J” to the “E” phase. A PIP joint progressing from the “E” to the “R” phase (“E/R” phase) is presented on image 32. Both IP joints show variable proportions of subchondral plate and synovial joint space remaining. The images 33 to 40 show fully remodeled “R” PIP joints with complete recovery of the joint.
Subchondral bone. Plate 5 exhibits 10 PIP finger joints with increasing proportions of normal subchondral bone left or reconstructed.
The top row shows the proximal (and distal) IP joints of a single digit with faint dotted vertical lines indicating the synovial joint space width.
The middle row shows the position of a rectangle square of which the height equalled the width of the joint space. The joint space was positioned in the centre of this square.
The images on the bottom row illustrate how the subchondral bone regions of interest were defined in the rectangular square. Areas where osteolytic activity and remodelling caused a disarrangement of the trabecular pattern, or where a complete loss of the trabecular structure had occurred are marked and dyed in these predefined region. Numbers indicate the proportional amount of subchondral bone with a normal trabecular structure in this region.
2