Correlation and diagnostic performance of metal ions in patients with pseudotumor after MoM hip arthroplasty: a systematic review and meta-analysis

Abstract

Background

The persistently rising complication, pseudotumor, after hip arthroplasty required surgeons’ vigilance. Although the remaining controversial relationship between metal ions and pseudotumor, metal ion detection had been widely employed in clinic. The aim of this study is to evaluate the correlation between metal ions and pseudotumor, as well as the effectiveness of metal ion analysis in the screening and diagnosis of pseudotumor through systematic review and meta-analysis.

Methods

The Medline and Embase databases were searched for studies evaluating metal ions and patients with pseudotumors after hip arthroplasty. A systematic review of risk ratio and diagnostic performance for metal ions was conducted.

Results

Seven studies were included in the systematic review. The mean Methodological Index for Non-Randomized Studies (MINORS) score of the included studies was 19 (range, 14 to 22). Pooled risk ratio (RR) value was 2.01(95% CI: 1.25–3.24; P = 0.004) for cobalt ions level and 1.44 (95% CI: 1.10–1.88; P = 0.008) for chromium ions level. The pooled sensitivity, specificity and the area under the curve (AUC) for cobalt and chromium ions were determined to be 0.59, 0.82, 0.73 and 0.34, 0.82, 0.56, respectively.

Conclusions

The metal ions level has a low diagnostic value. It is of certain value for confirmation, but should not be used as a routine screening indicator. The diagnostic value of cobalt ions is higher than that of chromium.

Level of Evidence: Diagnostic Level IV.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13018-024-05198-6.

Introduction

Total hip arthroplasty (THA) is the gold standard for the treatment of end-stage hip disease [1–3], while it has a number of complications [4, 5]. Among which, Pseudotumors are well-recognized complications in patients with metal-on-metal (MoM) THA, characterized by soft tissue reactions that can be cystic, mixed, or solid, comprising inflammatory cells and necrotic tissue [1, 6–9]. Pseudotumor, indicative of failed hip replacement, may result in pain, osseous denudation, pathological fractures, soft-tissue and muscle necrosis, hip dislocations, necessitating surgical revision [10]. recent studies have demonstrated pseudotumors in 10%-70% of patients after MoM hip arthroplasty, varying by population and implant type [11–13].

The metal ions released from the implants is currently considered as a main etiology of pseudotumor. One commonly accepted theory is that the interactions between metal species and body protein and cells generate a delayed type IV hypersensitivity reaction. Another acknowledged mechanism is that cell damage generated by metal ions may trigger the secretion of cytokines, leading to inflammatory reactions, then pseudotumors [14]. Therefore, the measurement of metal ions level is frequently utilized as an auxiliary diagnostic tool in the review of patients after metal hip replacement, as well as the prediction and screening for pseudotumors, despite ongoing debates regarding their roles in pseudotumor formation [15].

This study aims to systematically review and meta-analyze relevant research to assess the correlation between metal ion levels and the incidence of pseudotumors and to ascertain the utility of serum metal ion levels as a marker for screening and diagnosing pseudotumors.

Materials and methods

Literature search strategy

This study was conducted according to the PRISMA (Preferred Reporting Items for Systematic reviews and Meta-Analyses) statement. We conducted a comprehensive literature search using Medline and Embase database to determine the overall diagnostic performance of metal ions in detecting pseudotumor. The following search terms were used: ((Adverse Reactions to Metal Debris) OR (ARMD) OR (Adverse Local Tissue Reaction) OR (ALTR) OR (Granuloma, Plasma Cell) OR (Plasma Cell Granuloma) OR (Granulomas, Plasma Cell) OR (Plasma Cell Granulomas) OR (Inflammatory Pseudotumor) OR (Inflammatory Pseudotumors) OR (Pseudotumors, Inflammatory) OR (Pseudotumor, Inflammatory)) AND ((Arthroplasty, Replacement, Hip) OR (Arthroplasties, Replacement, Hip) OR (Arthroplasty, Hip Replacement) OR (Hip Replacement Arthroplasties) OR (Hip Prosthesis Implantation) OR (Hip Prosthesis Implantations) OR (Implantation, Hip Prosthesis) OR (Prosthesis Implantation, Hip) OR (Replacement Arthroplasties, Hip) OR (Replacement Arthroplasty, Hip) OR (Arthroplasties, Hip Replacement) OR (Hip Replacement Arthroplasty) OR (Hip Replacement, Total) OR (Replacement, Total Hip) OR (Total Hip Replacements) OR (Total Hip Replacement) OR (Total Hip Arthroplasty) OR (Arthroplasty, Total Hip) OR (Hip Arthroplasty, Total) OR (Total Hip Arthroplasties)) AND ((Ions) OR (cobalt) OR (Cobalt-59) OR (Chromium)). The search included studies published up until 20 march 2024, without any lower date limit.

Eligibility and exclusion criteria

Inclusion criteria were as follows: 1) studies in which pseudotumor developed after hip replacement and serum metal ions levels were measured; 2) usable data on clinical outcomes of the metal ions levels for pseudotumor patients after hip replacement (reporting the direct number of parts under and over the threshold: 5ppb (parts per billion, equivalent to 1 mg/L) [16–18] for each group, or to calculate them from the available data).

Studies were excluded: 1) used non-metallic hip prostheses; 2) lacked usable and computable data; 3) were low-quality studies; 4) were duplicately published; 5) were published in a language other than English.

Data extraction and quality assessment

Two reviewers (LHX and ZQY) participated in the data extraction procedure. The following data of each study were collected: 1) study characteristics, including: author, nation, publication year; 2) sample characteristics, including: sample size, average age, and proportion of women; 3) results of true positive (Tp), false positive (Fp), false negative (Fn), and true negative (Tn) (either directly report or calculated from the original published studies). Discrepancies were resolved through discussion, or negotiation with the third investigator. The risk of bias and methodological quality were evaluated using the MINORS checklist. Studies scoring thirteen or more points were included [19], any discrepancies were resolved through consensus.

Statistical analysis

We categorized patients into two groups based on metal ions concentration: those with > 5 ppb as the experimental group and those with ≤ 5 ppb as the control group, and compared the incidence of pseudotumors between them. The risk ratio RR and 95% confidence interval (CI) were calculated to to determine the relationship between metal ion concentration and pseudotumor occurrence.

Diagnostic metrics, including sensitivity, specificity, positive likelihood ratio (PLR), and negative likelihood ratio (NLR), were computed using standard formulas. The sensitivity of the tests was determined as Tp / (Tp + Fn), and specificity as Tn / (Tn + Fp). The PLR was calculated as sensitivity / (1 - specificity), and NLR as (1 - sensitivity) / specificity. Confidence intervals (95%) for these parameters were also computed. To calculate the area under the curve (AUC), the bivariate model was modified. Fagan’s nomogram was utilized to to estimate the likelihood of developing pseudotumors based on our diagnostic analysis. Heterogeneity among studies was assessed using the I² statistic, with a threshold of 50% to identify significant heterogeneity. Analyses were performed using a fixed-effects model when I² was below 50% and a random-effects model when above 50%. To explore potential sources of heterogeneity, we performed subgroup analyses to examine the effects of large-diameter MoM prosthesis, conventional MoM prosthesis, MoM resurfacing, and dual taper modular MoM prosthesis. Publication bias was evaluated using a funnel plot. Statistical analysis was performed using Stata 16.0 (StataCorp), with a two-sided P-value threshold of < 0.05 deemed statistically significant.

Results

Study identification

An initial search online turned in 632 studies from Medline and Embase. After subtracting 214 duplicates, 418 studies remained. Through the screening of titles and abstracts, 342 studies were removed. A full-text review of the remaining studies excluded 68 studies that 1) used non-metallic THA; 2) lacked usable and computable data; 3) were low-quality studies; 4) were duplicately published; 5) were published in a language other than English. The remaining 7 studies were included in this systematic review. The study’s identification, exclusion, and inclusion procedures were detailed in Fig. 1.

Fig. 1.

PRISMA flow diagram for literature selection process

Study characteristics

The demographic features and outcome parameters of the included studies are presented in Table 1. A total of 996 hip arthroplasty patients (365 patients with pseudotumor, 320 without pseudotumor, and 311 were not analysed) were included in this systematic review.

Table 1.

Characteristics and Outcomes of the Included Studies*

Nation	No. of Patients	Mean Age(yr)	Men/Women	Prosthesis Type	Test Sample	Colbat Ions†				Chromium Ions†
						Tp	Fp	Fn	Tn	Tp	Fp	Fn	Tn
Canada	372	54.3	262/110	Conventional MoM prosthesis	serum	21	15	12	13	10	3	23	25
American	92	54.7	46/46	Conventional MoM prosthesis	serum	30	24	19	19	28	22	21	21
American	148	58	99/49	Dual taper modular MoM prosthesis	serum	78	15	12	43	0	1	90	57
American	83	57.8	61/22	Dual taper modular MoM prosthesis	serum	22	0	8	53	0	0	30	53
American	156	65.2	69/87	MoM resurfacing	serum	24	4	83	45	13	2	94	47
The Netherlands	37	52.8	23/12	MoM resurfacing	serum	4	1	10	22	4	0	10	23
The Netherlands	108	60	52/56	Large-diameter MoM prosthesis	serum	27	35	15	31	26	37	16	29

^*MoM: metal-on-metal

^†The values are given directly or calculated from the literature

Methodological quality and publication bias

Seven studies were included in this study [16, 17, 20–24], the results of the risk-of-bias assessment are detailed in Table S1. The mean MINORS score was 19 (range, 14 to 22). The primary limitation affecting the assessed quality of these studies was the unblinded evaluation of objective and subjective endpoints. The studies by Chang et al. and Van der weegen et al. had the highest scientific quality. The Deeks test for funnel plot symmetry revealed the absence of significant publication bias (P _cobalt = 0.97; P _chromium = 0.32). Fig. S1 displayed the corresponding funnel plot.

Correlation analysis

The occurrence of pseudotumors was analyzed by categorizing patients into two groups based on metal ion concentration: > 5 ppb as the experimental group and ≤ 5 ppb as the control group. The forest plot revealed a pooled risk ratio (RR) of 2.01 (95% CI: 1.25–3.24; P = 0.004; I² = 88%) for cobalt ions and 1.44 (95% CI: 1.10–1.88; P = 0.008; I² = 56%) for chromium ions, indicating a significant association between increased metal ion levels and pseudotumor occurrence (Fig. S2).

Diagnostic performance

Analysis from seven studies on cobalt ions showed a pooled sensitivity of 0.59 (95% CI: 0.40–0.75) and specificity of 0.82 (95% CI: 0.53–0.95), with marked heterogeneity (I² > 90% for both metrics). The PLR, NLR, and AUC were calculated as 3.2 (95% CI: 1.1–9.5), 0.51 (95% CI: 0.32–0.80), and 0.73 (95% CI: 0.69–0.77) respectively (Fig. 2A and 3A). The pretest probability was determined to be 0.20, 0.44, and 0.60 according to the incidence of pseudotumor in patients after hip replacement, resulting in corresponding positive predictive values and negative predictive values of 44, 72, 83, and 11, 28, 43%, respectively (Fig. 4A).

Fig. 2.

Forest plot showed the pooled sensitivity and specificity of (A) cobalt ions level and (B) chromium ions level

Fig. 3.

Summary receiver operating characteristic curve (ROC) analysis of (A) cobalt and (B) chromium ions, the pooled ROC were 0.73 (95% CI: 0.69–0.77) and 0.56 (95% CI: 0.52–0.61), respectively

Fig. 4.

Fagan nomograms for the elucidation of post-test probabilities with different pretest probabilities. (A) Cobalt ions: The pretest probability was 0.20, yielding a PPV of 44% and a NPV of 11%; The pretest probability was 0.44, yielding a PPV of 72% and a NPV of 28%; The pretest probability was 0.60, yielding a PPV of 83% and a NPV of 28%. (B) Chromium ions: The pretest probability was 0.20, yielding a PPV of 39% and a NPV of 19%; The pretest probability was 0.45, yielding a PPV of 61% and a NPV of 39%; The pretest probability was 0.60, yielding a PPV of 80% and a NPV of 58%

For chromium ions, five studies were included into analysis [12, 13, 18–20]. The sensitivity was 0.34 (95% CI: 0.18–0.55) and specificity 0.82 (95% CI: 0.54–0.95). The PLR, NLR, and AUC were 2.0 (95% CI: 1.0–4.0), 0.80 (95% CI: 0.69–0.92), and 0.56 (95% CI: 0.52–0.61), showcasing substantial heterogeneity (I² > 90%) in these findings (Fig. 2B and 3B). The pretest probability was determined to be 0.20, 0.45, 0.60, and the corresponding positive predictive values and negative predictive values were 39, 61, 80, and 19, 39, 58%, respectively (Fig. 4B).

Subgroup analysis

To explore the sources of heterogeneity, meta-regression and subgroup analyses were conducted across different prosthesis types (group A: Large-diameter MoM prosthesis, group B: MoM resurfacing, group C: Dual taper modular MoM prosthesis, and group D: Conventional MoM prosthesis) as depicted in Fig. 5 for cobalt and chromium ions, respectively. MoM resurfacing was a particularly significant source of heterogeneity.

Fig. 5.

Univariable meta-regression and subgroup analysis for pseudotumor and different ions: (A) cobalt ions level; (B) chromium ions level

For cobalt ions, analyses revealed significant heterogeneity in sensitivity and specificity among the prostheses types (P < 0.05). Specifically, studies involving MoM resurfacing prostheses showed higher pooled sensitivity and specificity than those involving other prostheses types. Similarly, studies involving conventional MoM prostheses demonstrated notably higher specificities.

For chromium ions, MoM resurfacing was a significant source of heterogeneity in specificity (P < 0.05),which showing higher pooled specificities when compared to others.

Discussion

Pseudotumor was first recognized as a complication of MoM hip arthroplasty. While the role of metal ions in the development of pseudotumors was variably reported across both domestic and international studies [12, 16, 24–39]. Kwon et al. found that pseudotumors in patients with MoM hip resurfacing arthroplasties (MoMHRA) were significantly associated with elevated serum levels of cobalt and chromium ions [40]. Bosker et al. conformed the patients with elevated metal ion levels had a fourfold increased risk of developing a pseudotumor [16]. Conversely, Sutphen et al. found no statistical difference in pseudotumor formation with relation to elevated metal ions level [12], and Williams et al. were unable to discover a statistical association between the size of the pseudotumor and serum metal ion levels [34]. Furthermore, Bayley et al. found no correlation between the presence of pseudotumors and elevated serum metal ions level [28]. The relationship between different ions and pseudotumor development was also inconsistently described. Bisschop et al. demonstrated that an elevated blood cobalt level could predict pseudotumor formation, but denied the effect of chromium ions [35]. In contrast, Macnair et al. noted that the incidence of adverse reactions to metal debris (ARMD) was statistically higher with elevated blood chromium levels [15].

Despite the varied and often inconclusive studies on the relationship between metal ions and pseudotumor, the measurement of metal ion concentrations has become a well-established tool for the auxiliary diagnosis of pseudotumor, follow-up and risk assessment of pseudotumor in patients with MoM prostheses. Our meta-analysis identified a consistent association between elevated metal ion levels and the occurrence of pseudotumors, supporting some of the earlier findings. However, the elevated metal ions level had a weak diagnostic effect on pseudotumor detection.

This meta-analysis indicated that cobalt and chromium ions have good specificity (0.82) but poor sensitivity (0.59 for cobalt and 0.34 for chromium), with AUC of 0.73 and 0.56, respectively. The diagnostic accuracy of likelihood ratios (LRs): PLR and NLR, was highly valued in academic circles and clinical decision-making, surpassing sensitivity, specificity, or AUC. Ideally, for a diagnostic test to be considered highly useful, it should have a PLR exceeding 10.0 and an NLR less than 0.10 [41]. Proficiency in disease diagnosis is indicated by a greater PLR, while the disease exclusion is shown by a lower NLR. In this study, the PLR and NLR for cobalt and chromium in pseudotumor diagnosis were 3.2 and 0.51, and 2.0 and 0.80, respectively. These findings indicate that metal ions exhibit a poor diagnostic performance, though a comparatively greater ability to rule out pseudotumor diagnosis than to confirm it. Additionally, we used Fagan’s nomogram to leverage our diagnostic findings for estimating the likelihood of pseudotumor in routine clinical practice [42]. The results shown in Fig. 4 indicated that when evaluating a higher pretest probability, the confirming probability of pseudotumor was more suitable than excluding. Conversely, a lower pretest probability raised the probability of pseudotumor for excluding. When the pretest probability was noted to be 44% and 45% for cobalt and chromium ions, the likelihood of being pseudotumor free following a positive result was equal to the probability of having pseudotumor after a negative result. This suggested that metal ions performed comparably in terms of confirming and excluding pseudotumor. Thus the pretest probability of 44% and 45% can be the threshold of cobalt and chromium ions for diagnosing pseudotumor. If the anticipated pretest probability of pseudotumor was higher than the threshold, then the confirming strategy should be prioritized. While it was more appropriate to stress the excluding strategy when the predicted pretest chance of pseudotumor was below the threshold. These consisted with the poor agreement of the relationship between metal ions and pseudotumor described in literatures and the higher recognition of the association between cobalt ions and pseudotumor [20, 43–47].

This meta-analysis has certain limitations. The number of included studies was small. Only seven articles were included in the analysis, merely two were randomized controlled trials (RCTs). Due to the low number of relevant studies, and even less literature with available data. Besides, the exploration of heterogeneity was not comprehensive enough, and the influence of race, country, sample size, study type, etc. on heterogeneity was not examined. Additionally, the follow-up duration varied among the studies, the metal ions level and the incidence of pseudotumor may differ depending on the follow-up period.

Conclusion

This is the first meta-analysis to explore the relationship and diagnostic efficacy of metal ions level for hip pseudotumor. Pseudotumor is considered to be a low incidence postoperative complication. The incidence, which ranged from 10 to 70% according to recent literature and was comparable to typical surgical problems such dislocation, loosening, and periprosthetic infection, warranted the surgeons’ attention. This meta-analysis discovered an explicit connection between metal ions and hip pseudotumor. However, the metal ions level has a low diagnostic value. It is of certain value for confirmation, but shouldn’t be utilized for regular screening. Besides, the diagnostic value of cobalt ions is higher than that of chromium.

Author contribution

Concept/idea/research design: He-Xi Li; Acquisition of data: He-Xi Li; Analysis and interpretation of data: He-Xi Li; Writing/review/editing of manuscript: Hui-Qi Xie, He-Xi Li, Qing-Yi Zhang, Ning Sheng; Final approval of the manuscript: Hui-Qi Xie; Acquisition of funding: Hui-Qi Xie; Providing facilities/equipment: Hui-Qi Xie; Providing subjects: Hui-Qi Xie.

Funding

This work has been jointly supported by Sichuan Science and Technology Program (2024NSFSC0002); “1.3.5” Project for Disciplines of Excellence, West China Hospital, Sichuan University (ZYGD23037).

Availability of data and materials

No datasets were generated or analysed during the current study.

Declarations

Competing interest

The authors declare no competing interests.