Outcomes of search:
We searched 731 records up to December 2021 on ICH from electronic databases Cochrane Central Register of Controlled Trials (CENTRAL) (2), CINAHL (43), DOAJ (38), Embase (36), Medline (320), Proquest (3), Pubmed (238), and Scopus (51). We isolated 34 studies meeting the inclusion criteria from electronic databases (4-12, 14-27,29-39). Another study was isolated from the references of the isolated studies (40). We assessed 35 studies in qualitative analysis. There were 13 case reports, 20 case series, one clinical image, and a letter to the editor each. We excluded eight and one case from two duplicate studies due to similarity between the author names and study settings (41- 42). Another nine duplicate cases were excluded from a study because they were part of another study (19, 23). We excluded a case from the MRI review because of the insufficient description of ICH changes (43). We excluded two studies for operative interventions because their follow-up was less than six months (44-45). Figure 2 is the PRISMA diagram for the studies in this review. Annexure 2 details the reasons for exclusion of studies in PRISMA
Outcomes of Risk of bias assessment:
All studies included in this review were short case series/reports with a High Risk of bias. The Risk of bias assessment outcomes for this review is presented in Table 2. Thirty-five studies scored a mean of 4.82±2.02 points on the Risk of bias assessment tool in this review. 10, 14, and 11 studies rated poor, fair, and good on the Risk of Bias assessment tool.
We pooled 35 studies with 125 cases (136 hips) for this review. 11, 10, 9, 4, and 1 studies were from Europe, America, Asia, Africa, and Australia. The mean age of cases was 11.6 + 3.4 Years. There were 22 (18.6%) males and 96 (81.4%) females from 30 studies in the review. There were 43 left, 52 right, and 17 bilateral ICH hips. We pooled 46 cases/ 62 MRI reviews from 17 studies (4-10,12, 25-27, 34-37,39-40) for assessment of primary outcomes. More specifically, we had 46 cases/ hips (17 studies) where MRI was reviewed at least once. However, there were 14 follow-up MRI reviews from 8 studies among these cases (6-8, 10, 12, 25-27). Two studies added a second follow-up review in one case each (8, 25). We excluded one case from two studies due to insufficient description and poor MRI (9, 37). There were 8 cases/hips from 7 studies (7-10, 25, 35, 39) and 106 cases/115 hips from 22 studies (4-5,11-12,14-24, 26, 29-33,38) for assessment of secondary outcomes (pharmacological and operative interventions) for ICH.
Observations from the first MRI Reviews:
The Median time to first MRI after the presentation from 35 cases/15 studies was 1.75 (IQR 0.93-4.25) months. One study reported outcomes of MRI in 7 reviews at < 6 months, while another study presented outcomes of MRI reviews at a median of 2 (IQR 2-24) months (6, 26). The outcome measures on review of MRI characteristics from studies that were imaged at least once are summarized in Table 3. Though there were 43 observations of marrow edema (focal/diffuse) into the femoral head, acetabulum, or both, however, there was 26 focal geometric marrow edema into the femoral head with or without adjoining edema in the triradiate cartilage or the supra-acetabular region on MRI sequences among from 10 studies (4-5,7-9, 25, 27, 34-35, 40). The mean width of the geometric marrow edema reported from nine hips/3 studies was 8.88 (SD± 2.55) mm (4, 25, 35). One study reported the width of geometric edema in 2 cases >10mm (5). The remaining 17 cases which did not have focal marrow edema had a median time to the presentation of 6 (IQR 2-18) months from 7 studies (5-6, 10, 12, 26, 34, 36-37, 39). The two studies described before had all the cases without focal marrow edema (6, 26). We observed that the cartilage loss was inconsistently reported across studies. Seven studies performed cartilage-sensitive sequences to locate the site and the pattern of cartilage loss in this review Annexure 3 (5-6, 25-27, 36, 40). The pattern of cartilage loss was either focal or diffuse. The site of cartilage loss was acetabular or femoral head or both. Though the pattern of cartilage loss was documented, the site of loss was inconsistently reported. Seven studies comprising 26 hips (MRI reviews) did not comment on the pattern and location of cartilage loss, but 20 hips among them demonstrated joint narrowing due to cartilage loss (4-5, 7-9, 37, 39). On the other hand, two studies/hips quantified marked cartilage loss on MRI reviews (10, 36). Six studies with 17 cases had focal cartilage loss either centrally in the femoral head or superomedial weight-bearing area of the acetabulum (5-7, 26-27,40). Two studies/2 hips specifically reported no cartilage loss. However, both had joint space reduction (25, 35). We classified 25 hips/10 studies (4-5,7-8 25-27, 34-35,40) and 21 hips/ 10 studies (6, 9-10, 12, 26, 34, 36-37, 39) studies as early and late ICH in this review based on the observations of review authors.
Outcomes of Follow-up MRI reviews:
We had 14 follow-up MRI reviews from 8 studies among 46 cases (6-8,10, 12,25-27). The median time to follow-up MRI in these hips was 12.5 (IQR 3.75-19.5) months (6 studies). One study reviewed the follow-up MRI at a mean of 11 (9-16) months (4 hips) while another at less than 18 months (4 hips) (6, 26). The frequency of MRI characteristics among the follow-up MRI in these hips is given in Table 3. Eight hips demonstrated marrow edema (7 diffuse, one focal) while six did not. Four studies (4 hips) reported improved MRI characteristics, namely marrow edema and joint effusion on treatment (7-8, 10,25). These studies are discussed under the pharmacological interventions for ICH. However, four studies (10 hips) worsened in follow-up on MRI characteristics (6,12, 26-27). All except two hips which demonstrated MRI characteristics of worsening, had diffuse marrow edema or no marrow edema, to begin with.
We compared the change in the proportion of characteristics over time between the first and subsequent MRI reviews (Table 3). We observed a transition pattern from early to late changes between the MRI reviews. While early changes of geometric marrow edema and focal cartilage loss decreased significantly, the late degenerative changes into the femoral head and acetabulum, diffuse cartilage loss, and bone remodeling increased significantly between MRI reviews. We investigated the relationship between the presence of each MRI characteristic with time for the first MRI reviews.
(Figure 3). We noted that geometric marrow edema (P < 0.01), diffuse marrow edema (P < 0.05), diffuse cartilage loss (P < 0.05), and joint effusion (P < 0.05), were significantly associated across time of first MRI reviews.
We divided ICH treatment into pharmacological and Operative interventions groups.
Pharmacological interventions group (Table 4): We pooled 8 cases from 7 studies of ICH under the pharmacological interventions group . These studies were summarized into Etanercept (8, 10, 39), Methotrexate (7, 25, 35), and Botulinum neurotoxin toxin (9) sub-groups. We noticed that the included studies in the subgroup of pharmacological interventions were heterogeneous and used mixed methods to manage ICH. All except one study (8) started NSAIDs as the first treatment. Two studies used corticosteroids besides principal treatment (10, 35). Four studies performed additional procedures, including biopsy, joint debridement (open and arthroscopic), muscle releases, and bumpectomy (1, 7, 35, 39). All except one study started pharmacological treatment within a year of diagnosis (39).
Operative interventions for ICH (Table 5):We pooled 106 cases/115 hips from 22 reports under operative interventions group for ICH. There were 45 subtotal/partial capsulectomies ( 4-5, 14, 18, 21, 26, 29, 38), 18 THA (4- 5, 12, 18, 22), 14 open arthrotomies with or without biopsy and debridement(14, 16-17, 19-20, 24, 33), five arthrodiastasis ( 5, 15, 32) , five arthrodesis (18-19, 21, 23,) three osteotomies (16, 19, 29), three excision arthroplasties (19, 21), one each hanging hip operation, mould, cup and resurfacing arthroplasties (14, 19, 21). Tenotomies of muscle and soft tissue releases were either isolated or combined with other surgical procedures in 25 hips (14,17-18,22, 26, 29-30). Arthroscopy with or without associated procedures of labral debridement, abrasion chondroplasty, cartilage, and joint debridement with or without synovectomy was done for seven hips (11, 29, 31).