Comparison of Postoperative Serum Biomarkers after Total Hip Arthroplasty through Minimally Invasive Versus Conventional Approaches: A Systematic Review and Meta-Analysis of Randomized Controlled Trials

: Background: An effective way to objectively assess intraoperative tissue damage in total hip arthroplasty (THA) is to determine and compare postoperative serum biomarkers (laboratory parameters) such as creatine kinase (CK), C-reactive protein (CRP), and hemoglobin (Hb). This meta-analysis aims to compare the intraoperative tissue damage in THA through minimally invasive (MI) and conventional approaches (CAs) using postoperative serum biomarkers. Methods: We searched databases for randomized controlled trials (RCTs) comparing MI THA and CA THA. We calculated mean differences (MDs) with 95% conﬁdence intervals (CIs) for continuous outcomes, using the Hartung–Knapp–Sidik–Jonkman method and a common effect/random effects model. Results: A total of 13 RCTs, involving 1186 THA patients, were included in our meta-analysis. In two out of eleven examined outcome parameters, MI THA showed better results than CA THA. In nine out of eleven examined outcome parameters, MI THA showed no signiﬁcant difference compared to CA THA. MI THA had a 16 mg/L lower CRP value 3 days postoperatively than CA THA (I 2 = 66%, p = 0.03, MD = − 15.65, 95% CI − 30.10 to − 1.21). MI THA had a 3 mg/L lower CRP value 4 days postoperatively than CA THA (I 2 = 0%, p = 0.98, MD = − 3.00, 95% CI − 3.27 to − 2.74). Conclusions: Overall, there was no signiﬁcant difference between MI THA and CA THA in terms of postoperative serum biomarkers, with a slight advantage of MI THA in CRP values. These results do not provide sufﬁcient evidence to recommend changing the surgical approach from CA THA to MI THA. Level of evidence I: a systematic review of all relevant randomized controlled trials.


Introduction
Total hip arthroplasty (THA) is a promising solution for the treatment of many hip conditions such as osteoarthritis, femoral neck fracture, dysplasia, and avascular necrosis of the femoral head (ANFH) [1][2][3]. According to the anatomical relationship to the greater trochanter, there are six surgical approaches to the hip joint: anterior, anterolateral, direct lateral (transgluteal or transtrochanteric), posterior, posterolateral, and superior [4]. In an attempt to improve treatment outcomes, minimally invasive (MI) THA has been introduced and further developed over the past two decades. MI surgical approaches to the hip joint are modifications of the well-known conventional approaches (CAs) that must meet two conditions: an incision length ≤ 10 cm and, most importantly, the preservation of muscles and tendons. The assumption that MI approaches would lead to a significantly better patient outcome due to less tissue damage is something that still needs to be scientifically proven [5][6][7][8][9]. A simple and effective way to objectively assess intraoperative tissue damage is to determine and compare postoperative serum biomarkers (laboratory parameters) such as creatine kinase (CK), C-reactive protein (CRP), and hemoglobin (Hb). CK is one of the important blood proteins that are produced by the breakdown of muscle fibers. It, therefore, serves as one of the most commonly used indirect markers of intraoperative muscle damage in humans [10]. CRP is another blood protein that is produced in the liver. It is the most important blood laboratory value for detecting and monitoring inflammation in humans [11]. The connection between tissue damage and inflammation is well-known [12], which is why CRP can be used to indirectly assess intraoperative tissue damage [11]. Another way to estimate intraoperative tissue damage is to measure the blood loss. The overall blood loss can be determined very reliably using the hemoglobin value, since its determination allows conclusions to be drawn about the hidden blood loss, whereas measuring the amount of intraoperative blood loss and the postoperative drainage directly only reveals a share of the overall blood loss [13]. Despite the long-standing use of MI approaches in THA, there is no meta-analysis in the literature examining differences in postoperative serum biomarkers between MI and CAs. One recently published systematic review of the literature by Sarantis et al. provided important findings on this topic, but without performing a meta-analysis of the extracted data [14]. There is a need for an objective assessment of the extent of intraoperative tissue damage in THA between MI approaches and CAs using postoperative serum biomarkers. The aim of this study is to confirm or refute the hypothesis that the extent of intraoperative tissue damage in THA through MI approaches, as measured by postoperative serum biomarkers, is less than in THA through CAs.
We formulated the following PICO (Population, Intervention, Control, and Outcomes) question: In human participants with hip conditions such as osteoarthritis, femoral neck fracture, dysplasia, and ANFH is MI THA superior to CA THA in terms of postoperative serum markers (CK, CRP, and Hb)?

Data Sources and Search Strategies
We registered our study protocol in PROSPERO on 10 August 2022 (CRD42022350279). We followed the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) guidelines [15]. The PRISMA checklist is available in Table S1 in the supplementary materials. The author group of the present study has certain experience in the field of meta-analyses and THAs. For this reason, the methods described in some publications [16][17][18][19] of this author group are similar or partially identical to the present meta-analysis. We searched the following databases and checked citations of the related meta-analyses for relevant manuscripts up to 31 March 2023: PubMed, CNKI, The Cochrane Library, Clinical Trials, CINAHL, and Embase. We built a BOOLEAN search strategy, adapted to the syntax of the used databases. No restrictions to publication language were applied. Since there are constant advances in THA, especially for MI surgical techniques, we decided not to include old studies published before 2010.

Study Screening and Selection
First, we examined the titles, then the abstracts, and finally the full texts of the articles. The decision on the inclusion of each study was made by the consensus between two reviewers (NR, PMK). We used the Kappa coefficient to measure the agreement between them. Disagreements were resolved by scientific discussion.

Inclusion Criteria
Types of studies: • randomized controlled trials (RCTs).
Types of participants: • human participants with hip conditions such as osteoarthritis, femoral neck fracture, dysplasia, and ANFH.

Types of interventions:
• MI THA or CA THA.
MI THA definition: In our meta-analysis, an approach was defined as MI if the approach per se is known as MI in the literature or if the approach was explicitly referred to as MI in the individual RCTs.

Data Extraction and Quality Assessment
Two reviewers (NR, PMK) independently extracted all relevant data on RCT characteristics, methods, quality assessment, characteristics of participants, details of the interventions, relevant outcomes, and relevant additional information. Disagreements were resolved by scientific discussion. The extracted data are available in Excel S1 in the supplementary materials. We performed a risk of bias assessment according to Cochrane's Risk of Bias 2 (RoB 2) tool [20] and a level of evidence assessment according to the recommendations of the GRADE system [21]. In addition, we assessed publication bias using Begg's and Egger's tests.
The RoB 2 tool can be used to estimate the likelihood that study design features cause misleading results [20]. After considering all quality aspects, an overall rating of "low risk", "moderate risk", or "high risk" is made for each study [20]. The results of a "low risk" study are considered valid, while the results of a "high risk" study might be considered invalid [20].
GRADE is a tool that allows assessment of the quality of evidence, using four levels of evidence: "very low", "low", "moderate", and "high" [21]. The evidence levels indicate to what extent the true effect deviates from the estimated effect [21]. Since the quality of the evidence often varies between the outcome parameters, it is determined individually for each endpoint [21].
The publication bias indicates a statistical distortion of the data presentation in scientific journals, which can occur as a result of a preferred publication of studies with "positive" or significant results.

Measures of Treatment Effect
We calculated mean differences (MDs) with 95% confidence intervals (CIs) for continuous outcomes, using the Hartung-Knapp-Sidik-Jonkman method and a common effect/random effects model. The Hartung-Knapp-Sidik-Jonkman method is a simple and robust approach to meta-analysis that is gaining more popularity among statisticians. It far surpasses the standard method of DerSimonian-Laird [22]. Common effect (or fixed effects) and random effects models are statistical models, both of which are regularly used in meta-analyses. A common effect model assumes only one true effect, which can be disadvantageous if there is considerable heterogeneity between the primary studies. A random effects model assumes that the actual effect may vary due to heterogeneity within the studies examined [23]. We performed study weighting by inverse variance. We calculated the t-test to determine the differences between the means of the two groups. We assessed heterogeneity using Cochrane's Q test (p-value < 10 is indicative of heterogeneity) and Higgins test I 2 (low heterogeneity: <25%, moderate heterogeneity: 25-75%, and high heterogeneity: >75%) [24]. As these values indicated a high amount of heterogeneity for some parameters, we adhered to the random effects model in our result presentation. Taking measurement accuracy into account, the results are reported with two decimal places. All statistic calculations were performed using the R packages meta and metafor.

Missing Data
We contacted the authors of the RCTs for missing data. If relevant data were still missing, the RCT was excluded to ensure the high-quality inclusion of RCTs. If the information on standard deviation (SD) was missing, it was calculated via imputation [25]. In case the RCTs provided different information on intention to treat (ITT) and per protocol (PP) analysis, we used the numbers from the ITT analysis.

Hb 3 Days Postoperatively: MI THA vs. CA THA
Data on 293 THAs were pooled from three RCTs (I 2 = 85%, p < 0.01, Figure 12). The Hb 2 days postoperatively of MI THA showed no significant difference compared to the Hb 3 days postoperatively of CA THA (MD = 0.01, 95% CI −1.22 to 1.23).

Discussion
Our literature search revealed 13 RCTs with 1186 THA patients which were included in the final meta-analysis. The extent of the included RCTs allowed us to examine three relevant postoperative serum biomarkers (CK, CRP, and Hb). In two out of eleven examined outcome parameters, MI THA showed significantly be er results than CA THA. In nine out of eleven examined outcome parameters, MI THA showed no significant difference compared to CA THA. Taking the examined outcome parameters into account, we allow ourselves to state that MI THA and CA THA show overall no significant difference in terms of postoperative serum biomarkers, with a slight advantage of MI THA in CRP values. Therefore, these differences do not justify changing the operative approach based on the examined outcome parameters. The choice of the operative approach still should be left to the experience and preference of the operating surgeon.
We examined CK and CRP on postoperative days 1-4 and Hb on postoperative days 1-3. The combination of the 11 examined outcome parameters provides an objective impression of the extent of intraoperative tissue damage in THA. Figures 2-12 show that MI THA had statistically be er results or no significant difference in postoperative serum biomarkers compared to CA THA. MI THA had a 16 mg/L lower CRP value 3 days postoperatively than CA THA. MI THA had a 3 mg/L lower CRP value 4 days postoperatively than CA THA. When interpreting statistically significant differences, the question of clinical relevance is crucial. Minimal clinically important differences (MCIDs) are patient-derived scores that justify changes in a clinical intervention. Although the literature does not provide any information on specific values for the examined postoperative serum biomarkers (CK, CRP, and Hb) in THA, the statistically significant differences do not seem to reach MCID at first appearance.
Our meta-results are similar to the findings of the recently published systematic review on this topic [14]. The study of Sarantis et al., including 31 studies, found no relevant differences between THA approaches when evaluating biomarkers (CK, CRP, myoglobin, erythrocyte sedimentation rate, skeletal troponin, and interleukins) [14]. Only a slight advantage of anterior and minimally invasive approaches was noted [14].
There are some strengths of this meta-analysis that need to be highlighted. To the best of our knowledge, this is the first systematic review and meta-analysis examining differences in postoperative serum biomarkers between MI THA and CA THA. In our meta-analysis, we applied high-quality statistical methods. The literature search was limited to RCTs to obtain more reliable meta-results. The RCT quality was determined by assessment of the risk of bias, the level of evidence, and the publication bias. Two statistical models, namely the common effect and the random effects models, were calculated. Furthermore, the Hartung-Knapp-Sidik-Jonkman method was used instead of the standard DerSimonian-Laird method, which is recommended in the recent literature [22].

Discussion
Our literature search revealed 13 RCTs with 1186 THA patients which were included in the final meta-analysis. The extent of the included RCTs allowed us to examine three relevant postoperative serum biomarkers (CK, CRP, and Hb). In two out of eleven examined outcome parameters, MI THA showed significantly better results than CA THA. In nine out of eleven examined outcome parameters, MI THA showed no significant difference compared to CA THA. Taking the examined outcome parameters into account, we allow ourselves to state that MI THA and CA THA show overall no significant difference in terms of postoperative serum biomarkers, with a slight advantage of MI THA in CRP values. Therefore, these differences do not justify changing the operative approach based on the examined outcome parameters. The choice of the operative approach still should be left to the experience and preference of the operating surgeon.
We examined CK and CRP on postoperative days 1-4 and Hb on postoperative days 1-3. The combination of the 11 examined outcome parameters provides an objective impression of the extent of intraoperative tissue damage in THA. Figures 2-12 show that MI THA had statistically better results or no significant difference in postoperative serum biomarkers compared to CA THA. MI THA had a 16 mg/L lower CRP value 3 days postoperatively than CA THA. MI THA had a 3 mg/L lower CRP value 4 days postoperatively than CA THA. When interpreting statistically significant differences, the question of clinical relevance is crucial. Minimal clinically important differences (MCIDs) are patient-derived scores that justify changes in a clinical intervention. Although the literature does not provide any information on specific values for the examined postoperative serum biomarkers (CK, CRP, and Hb) in THA, the statistically significant differences do not seem to reach MCID at first appearance.
Our meta-results are similar to the findings of the recently published systematic review on this topic [14]. The study of Sarantis et al., including 31 studies, found no relevant differences between THA approaches when evaluating biomarkers (CK, CRP, myoglobin, erythrocyte sedimentation rate, skeletal troponin, and interleukins) [14]. Only a slight advantage of anterior and minimally invasive approaches was noted [14].
There are some strengths of this meta-analysis that need to be highlighted. To the best of our knowledge, this is the first systematic review and meta-analysis examining differences in postoperative serum biomarkers between MI THA and CA THA. In our meta-analysis, we applied high-quality statistical methods. The literature search was limited to RCTs to obtain more reliable meta-results. The RCT quality was determined by assessment of the risk of bias, the level of evidence, and the publication bias. Two statistical models, namely the common effect and the random effects models, were calculated. Fur-thermore, the Hartung-Knapp-Sidik-Jonkman method was used instead of the standard DerSimonian-Laird method, which is recommended in the recent literature [22].
In day-by-day clinical work, the main findings of this meta-analysis give important insight for orthopedic surgeons. The knowledge that the postoperative serum biomarkers of MI THA and CA THA do not differ relevantly could help orthopedic surgeons choose their surgical approach and operational technique. Furthermore, the calculated weighted mean values of the postoperative serum biomarkers CK, CRP, and Hb represent a reliable reference value that can be compared to the postoperative laboratory values after THA in other hospitals for critical self-control.
We identified the following limitations to our meta-analysis. (1) Significant heterogeneity was detected between the included RCTs for several outcome parameters. (2) The included studies summarized different surgical indications in a meta-analysis: osteoarthritis, femoral neck fracture, dysplasia, and ANFH. (3) In some outcome parameters, the sample size and the number of RCTs included were small. (4) The follow-up period of 4 days in our meta-analysis is relatively short. However, the short follow-up period results from the reported data that were obtained from the primary studies. The investigation of postoperative serum biomarkers with a longer follow-up period would be desirable in future studies. (5) A systemic increase in postoperative serum biomarkers such as CK, CRP, and Hb can be attributed to factors other than the THA approach. Such confounding factors might be infections or other inflammatory reactions, extensive postoperative rehabilitation, renal insufficiency, and adverse drug reactions.

Conclusions
Our meta-analysis indicates that there was no significant overall difference between MI THA and CA THA in terms of postoperative serum biomarkers (CK, CRP, and Hb). We found a slight advantage of MI THA in CRP values. MI THA had a 16 mg/L lower CRP value 3 days postoperatively than CA THA. MI THA had a 3 mg/L lower CRP value 4 days postoperatively than CA THA. However, these findings do not provide sufficient evidence to recommend changing the surgical approach from CA THA to MI THA, since the differences between the examined approaches did not seem to reach MCID.