Background: Femoroacetabular impingement (FAI) is a highly prevalent painful condition that is considered a major risk factor for the early development of hip osteoarthritis (HOA). The molecular pathway behind the association between FAI and HOA remains unclear. Unraveling the elements that constitute this pathway is a milestone in the search of innovative methods to prevent HOA - not only for patients with FAI.

Objectives: This systematic review with meta-analysis aimed to assess the differences in the molecular expression of hip joint tissues between patients with i) FAI without HOA; ii) HOA due to FAI; and iii) healthy hip joints.

Methods: Our study (PROSPERO: CRD42023404290) was conducted following the PRISMA statement. Two independent reviewers performed the study selection and data extraction steps. Embase, CINAHL, LILACS, PubMed, SCOPUS, SPORTDiscuss and The Cochrane Library were screened. We searched for all primary studies assessing the differences in the molecular expression of hip joint tissues of patients with early-staged FAI (without HOA) compared to patients with HOA due to late-staged FAI, with or without comparisons with healthy joints. The risk of bias (RoB) of these studies was assessed by using a checklist based on the “criteria rated by five tools” [1]. Studies presenting low or moderate RoB were selected for meta-analysis, in which we used a random-effects model. We considered the mean and standard deviation of normalized units [for each element being compared in more than 1 study] to perform meta-analyses. For comparisons across 4 different studies or more, we calculated the standardized mean difference (SMD; 95% CI). Otherwise, we calculated MD.

Results: Nineteen studies were selected for full-text reading, of which 9 were included for RoB assessment. Three studies were excluded due to a high RoB. Inclusion of patients with hip dysplasia in the healthy group and insufficient description regarding methods and sample characteristics were the main sources of RoB. For the remaining 6 studies [2-7], a summary can be found in the Figure 1. Femoral chondral expression (FCE) of IL-1β was higher in FAI compared to healthy joints [MD 40 (33-47); I² = 1%]. Compared to HOA, FCE of IL-1β in FAI was higher in a study [2] and similar in another [3]. Synovial expression of IL-1β was higher in HOA compared to FAI [MD 30(24-37); I²=70%]. FCE of MMP-13 and ADAMTS-4 was higher in FAI compared to healthy joints [SMD 5(2-8); I²=86%]; [MD 38(33-43); I² = 1%] - respectively. Compared to HOA, FCE of MMP-13 in FAI was similar in 3 studies [3, 4, 7], higher in 1 study [2], and lower in another [5]. FCE of DNMT3B was higher in healthy joints compared to FAI [MD 8(5-11); I² = 1%], and higher in FAI compared to HOA [MD 4(3-6); I² = 1%]. Labral expressions of IL-1β, IL-8, MMP-3 and COL1A1 were significantly higher in HOA compared to FAI. FCE of PPARγ decreased with disease progression (i.e., healthy > FAI > HOA) [7], while FCE of DNMT1, DNMT3A [7] and ABAT [4] increased with disease progression. Compared to healthy joints, FCE of FasL, p21, Bcl2 [6], and NITEGE [3] in FAI was significantly higher. Synovial expression of COL1A1, IL-8 and MMP-3 was significantly higher in HOA compared to FAI.

Conclusion: The hip chondral microenvironment in FAI has a predominance of pro-inflammatory chondrolysis-related (IL-1β, MMP-13, ADAMTS-4, NITEGE) and senescence-related (FasL, p21, Bcl2) expression compared to healthy joints, and is often similar to that found in HOA. Labral and synovial microenvironments in HOA have a predominance of expression associated to extracellular matrix degeneration (COL1A1, MMP-3) and inflammation triggering (IL-8, IL-1β) compared to FAI. Perhaps, the key for progression from FAI to HOA lies in extra-chondral joint tissues. The decrease in the hip chondral expression of DNMT3B and PPARγ with disease progression (healthy > FAI > HOA) indicates a loss of control on chondrocytic apoptosis and on extracellular matrix regeneration in FAI and, even worse, in HOA.

REFERENCES: [1] PMID: 33964880.

[2] PMID: 26791034 (Chinzei).

[3] PMID: 32736506 (Haneda).

[4] PMID: 36606425 (Kamenaga).

[5] PMID: 36463522 (Kuhns).

[6] DOI: 10.1016/j.jcjp.2021.100011 (Liang).

[7] PMID: 35288246 (Pascual-Garrido).

Acknowledgements: NIL.

Disclosure of Interests: None declared.

• Download figure
• Open in new tab
• Download powerpoint