Background/Purpose: Magnetic resonance imaging (MRI) is more sensitive than physical examination in detecting inflammation. This has predominantly been studied in metacarpophalangeal-(MCP) and wrist-joints. Data on the concordance and discordance of physical examination (PE) and MRI-detected inflammation of metatarsophalangeal-(MTP)joints is scarce, which is  surprising as physical examination of  these joints is challenging.We aimed to study the concordance and discordance of arthritis upon PE with MRI-detected inflammation of MTP-joints. Analyses on MCP-joints were included for comparison.

Methods: 1764 MTP(2-5)-joints and 1764 MCP(2-5)-joints of 441 consecutive patients presenting with early inflammatory arthritis (36% RA, 64% other inflammatory arthritides) underwent PE of joint swelling by two independent readers, and 1.5T contrast-enhanced MRI of unilateral MTP- and MCP-joints. MRI-detected synovitis and bone marrow oedema were scored according to the RA MRI score (RAMRIS), and tenosynovitis according to Haavardsholm by two experienced readers. For a joint to be classified as PE+ swelling had to be observed by both assessors. MRI-positivity required the presence of the same MRI-inflammatory feature on joint level that was scored by both readers≥1. MRIs were further studied by two other, independent observers among whom an experienced musculoskeletal radiologist by consensus reading, to investigate the presence of contrast-enhancement of the soft tissues that was not scored according to RAMRIS guidelines. Analyses were done on joint level, joints were grouped as PE+MRI+, PE–MRI–, PE+MRI– and PE–MRI+.

Results: Physical examination of joints and MRI were concordant in 79% of MTP-joints (4% PE+MRI+, 74% PE–MRI–). For the MCP-joints this was 71% (15% and 53% respectively). Next discordance was studied. Subclinical joint inflammation (PE–MRI+) was present in 17% of MTP-joints. This was less frequent than in MCP-joints, where subclinical inflammation was present in 29% joints. Discordance in the opposite direction (PE+MRI–) was present in 5% of all MTP-joints (this was 55% of all PE+ MTPs). PE+MRI- occured less frequent at MCP-joints (3% of all MCPs were PE+MRI, which is 15% of the PE+ MCPs). Subsequently, MRI-detected contrast-enhancement of the soft tissues were studied to see if they could form an explanation for the felt arthritis. Of the PE+MRI- MTP-joints, 43% showed contrast-enhancement of the soft tissues, while in PE-MRI- joints this was 22%. Finally, to better understand this contrast-enhancement, its’ frequency was compared between RA and other inflammatory arthritides, And was 53% vs. 22% respectively (OR 4.1, 95% CI 3.3-5.0).

Conclusion: Joint examination and MRI were mostly concordant in MTP- and MCP-joints. In MTP-joints MRI-detected subclinical joint inflammation occured in 14% of joints, which was less frequent than the MCP-joints where this was 27%. Clinical joint swelling without MRI-detected joint inflammation according to RAMRIS occured more in the MTPs and was in part caused by contrast-enhancement of soft tissues; this was more frequent in RA suggesting it to be inflammatory in nature. Further detailed imaging studies are needed to further characterize its‘ nature

ACR Meeting Abstracts - https://acrabstracts.org/abstract/how-accurate-is-physical-joint-inflammation-of-the-mtp-joints-and-what-can-we-learn-from-additional-mri-on-forefoot-involvement-in-early-arthritis/