Safety and Ecacy of Hemispherical with Flattened Pole Dual-Mobility Acetabular Cup in Primary Hip Arthroplasty: The SYMCOR-1 study

Background: This study estimated the short-term clinical safety and ecacy of hemispherical with attened pole cobalt-chromium metal-back with porous outer hydroxyapatite coating dual-mobility acetabular cup (HFPC-DM-HA) in primary total hip arthroplasty. Methods: Single-center retrospective observational cohort study of consecutive patients undergoing total hip arthroplasty with a HFPC-DM-HA 2 years prior to study start. Prospective 2-year follow-up with letter and phone questionnaires. Results: Sampling frame: 361 patients including 59 patients (16.3%) in the cohort. 6 patients (10%) lost to follow-up. Median age 77.5 years (range: 67 ; 92), 32% female, median BMI 25.2 kg.m -2 (18.4 to 56.8). Clinical indications: Primary THA in all patients, resulting from primary osteoarthritis in 80% of them. Median follow-up 3.0 years (2.7 to 4.1). Primary endpoint: 2-year implant survival rate: 97% [87, 99]. Prosthetic dislocation: 0%. Secondary endpoint: Modied HHS (pain & functional subscore) improved from baseline 39.7 [34.6, 44.7] to 75.8 [72.1, 79.6] at 1-year and to 86.7 [83.7, 89.7] at 2-year follow-up (p<0.0001). Conclusions: The authors deemed the short-term outcomes of this acetabular cup in primary total hip arthroplasty to be satisfactory. Study

estimate short-term safety and e cacy of HFPC-DM-HA in primary THA in "real-life" practice, prior to considering a long-term prospective study.

Study design
This was a single-center retrospective observational cohort study of all consecutively operated patients who underwent THA with a HFPC-DM-HA prior to study start and who were eligible for a 2-year postoperative assessment at the time of the study.
This study was subject to MR3 regulation and was therefore noti ed to the CNIL commission without requiring medical ethics committee approval. The data source was the complete database of patient charts on March 1, 2018, including the sampling frame that included cohort patients.
Two-year follow-up status and missing information about cohort patients were obtained by mailed questionnaires and telephone interviews. The information letters were drafted according to regulations and informed patients that they may refuse participating in this study.

Patients
The investigator's operative records were screened between his rst SYMBOL DM HA implantation from July 3, 2014 through December 17, 2015 and an exhaustive list of THAs was established. The sampling frame consisted of all patients who had undergone THA during that period while the cohort was the subset of patients in whom the primary THA had been performed using HFPC-DM-HA. Anonymous data from patient charts were recorded into a database for the entire sampling frame and included demographics, operative date, whether the THA was a primary or revision surgery as well as acetabular cup model. Detailed preoperative, operative and postoperative data were recorded for the HFPC-DM-HA cohort only.
Patient inclusion criteria in the cohort were primary THA performed by the investigator during the screening period using a HFPC-DM-HA. Exclusion criteria were patient refusal to participate in the study, minors less than 18 years of age and patients under guardianship. No patient was excluded from the cohort because of the type of femoral stem, the need for additional surgery or missing data.
Standard patient charts at the investigation site included physical, functional and radiographic assessments preoperatively and at 1-year follow-up. Intermediate assessments between the rst and fth year follow-up were not common practice at that site, unless patients reported an adverse event or required surgery on another joint, so the 2-year follow-up of most patients consisted of self-reported outcomes recorded in a questionnaire that had been mailed to the patient or a telephone interview in case of missing or inconsistent information.

Intervention
The index operation was past primary THA on the target hip using a HFPC-DM-HA. Cup xation was press-t only. All femoral heads used were CoCr. The surgeon used the stem deemed the most suitable on an individual patient basis. Additional surgery was performed if required.

Endpoints
The primary endpoint was cup survival up to two years post-implantation. The endpoint was de ned as joint patient survival and non-removal of the cup.
The secondary safety endpoints were: The rate of intraoperative adverse events and the rates of postoperative implant-related or procedure-related post-operative adverse events over 2-year follow-up. The rates of prosthetic dislocation and intraprosthetic dislocation (IPD) were analyzed. IPD was de ned as the femoral head dissociating from the mobile bearing PE liner. [1,2] The secondary effectiveness endpoints were the Harris Hip Score (HHS) and the modi ed HHS (mHHS) that consisted in the sum of pain & functional subscores without the range of motion and deformation. While the HHS could be computed preoperatively and at 1-year follow-up, the 2-year follow-up questionnaire only enabled computing the mHHS.

Statistical analysis
Descriptive statistical analysis of the sampling frame was performed on gender, age at the time of surgery, primary vs. revision THA and acetabular cup type. The HFPC-DM-HA cohort was compared to its sampling frame with respect to those variables.
The cohort's demographic, preoperative, operative and postoperative descriptive statistical analysis was performed. Adverse events were tabulated and counted. Implant survival was analyzed using the Kaplan-Meier survivor function. [3] The means of quantitative variables were compared between groups using the unpaired t-test when applicability criteria were met. [4,5] The two-sample Wilcoxon-Mann-Whitney ranksum non-parametric test was used otherwise. Mean changes in scores within individuals were tested using the paired t-test when applicable and the Wilcoxon signed-rank test was used otherwise. [6][7][8] Frequencies of categorical variables between independent groups were compared using the Chi-square when applicability criteria were met and the Fisher exact test was used otherwise. [9][10] Binomial categorical variables equality to 0.5 was tested using the exact binomial probability test. Ordinary leastsquares linear regression of overall HHS vs. mHHS were plotted using preoperative and 1-year follow-up data in order to estimate how closely 2-year the mHHS could predict the 2-year HHS. [11] The analysis was conducted on complete cases providing missing data did not exceed 10%. Statistical analyses were conducted with a script programmed in STATA 15 software (StataCorp, College Station, TX, USA).

Patient disposition
The sampling frame consisted of 361 patients who had undergone primary THA. Fifty-nine cases used HFPC-DM-HA and all of them were included in the cohort. Six patients (10%) were lost to 2-year follow-up (Fig. 2).
All HFPC-DM-HA were press-t without the use of cement, screws or bone grafting. All had a PE liner tted with a CoCr alloy femoral head. Acetabular shell diameters ranged from 48 to 56 mm and the most frequent used diameter range (54%) was 50 to 54 mm. Femoral stems were all cementless (100%).
One patient required intra-operative additional procedure for shell repositioning after testing.  (Fig. 3). That rate was calculated over the initial cohort of 59 patients taking into account the 2 known failures, 1 death and the 7 patients lost to follow-up.

Secondary endpoints: Postoperative implant or procedurerelated complications
Post-operative adverse events, whether related or not to the procedure or implant, were reported in 39 patients (66%) presented (Table 3). Nineteen patients (32%) presented postoperative complications that were adjudicated as procedure and/or implant related. These included the 2 events that caused prosthetic hip revision described previously, 13 with pain in the ipsilateral hip, pelvis or thigh and 4 with surgical scar-related problems. No prosthetic dislocation including IPD was reported during follow-up. Secondary endpoints: Functional outcomes HHS and mHHS were closely correlated preoperatively (R-squared: 0.9993, p < 0.0001,) and at 1-year follow-up (R-squared: 0.9995, p < 0.0001, Fig. 4). Detailed analysis of scores showed that the change in HHS was mostly driven by the change in mHHS items. This suggested that the mHHS at 2-year follow-up could be reasonably compared to prior measurements, although the range of motion and deformation subscores were not available at 2-year follow-up. Pre-and postoperative HHS and mHHS summary in Table 3 Mean within-patient HHS increased from preoperative baseline to

Discussion
Need for this study The safety and e cacy of medical devices are functions of several critical technical characteristics and the interplay between those characteristics. For that reason, the clinical risk-bene t of an implant with a given combination of critical characteristics cannot be predicted by examining the risk-bene t related to each characteristic separately reported in other models with different combinations of the critical characteristics. The European medical device clinical evaluation guideline requires device-speci c clinical safety and performance data to be presented in order to establish the bene t-risk balance of medical device with a speci c combination of critical characteristics. [14] That requirement was reinforced with the introduction of the European Medical Devices Regulation. [15] Predicting the bene t-risk balance of a new medical device based on clinical evidence derived from a previously approved "predicate" device, is valid only if the two devices meet equivalence criteria and requires the same combination of critical characteristics and the same intended use. In the case of DM, equivalence requires shells to share the same combination of metal-back design and alloy, coating, xation mechanism, clinical indications and any other feature that could modify clinical outcomes. This study was conducted because a systematic review of published clinical studies with DM revealed that HFPC-DM-HA had no predicate device. That systematic review was beyond the scope of this article, but shell differences were shown with a broad range of DM with clinical evidence reported in a compilation of articles (Table 4). [16].

Internal validity
The internal validity of this study is ensured by consecutive recruitment all eligible cases performed by a single surgeon and by their systematic follow-up. The limitations were the relatively small sample size, the relatively short follow-up duration, the inability to perform systematic physical and radiographic assessments at 2-year follow-up, and the lost to follow-up. Patient contacts along with information retrieved in patient charts suggested that missingness was not procedure-related or implant related. The strong associations between the HHS and the mHHS preoperatively, at 1-year follow-up and between their changes within patients over those two periods suggest that mHHS at 2-year follow-up is a reasonable estimator of the e cacy of THA with HFPC-DM-HA at that speci c period, in the absence of 2-year physical and radiographic assessments to calculate the HHS. A secondary analysis of implant survival after multiple imputations of missing data was considered, but given the entire dataset at 2-year followup was missing in the concerned patients that method was not implemented.

External validity
The external validity of this cohort study was based on the demonstrated completeness of recruitment and comparison with the sampling frame. The main limitations were the single-center recruitment and small sample size. It would be tempting to compare the results of SYMCOR-1 with those of a prospective "Regular Dual Mobility" cohort study of patients with similar demographics followed up 2 to 5 years. [17] However, this would require adjusting for differences in clinical indications and xation mechanism.
Pooling the results of studies on DM and comparing them should be performed in the course of a rigorous systematic review with risk of bias assessment and meta-analysis. Such a process would focus on common outcome variables and stratify over clinical indications, surgical/implant differences and risk factors.

Conclusion
This was the rst cohort study to present two-year follow-up safety and e cacy data on the use of a dual-mobility acetabular cup in primary total hip arthroplasty with a hemispherical with attened pole cobalt-chromium metal-back shell with outer porous double layer titanium/hydroxyapatite.

Consent for publication
This article does not contain individual data.

Availability of data and materials
The dataset generated during and/or analyzed during the current study are available at 1 Clinique de la Sauvegarde, Lyon, France. The datasets generated during and/or analysed during the current study are not publicly available due to individual data protection law but speci c anonymous subsets will available from the corresponding author on reasonable request.
Competing interests NB, GE, JG and OG are bene ciaries of royalties paid by the manufacturer of the study devices, who is the sponsor of this study.
FD is a consultant in biostatistics and clinical research appointed by the sponsor.

Funding
This work was sponsored and funded by Dedienne Santé S.A.S.
Author's contributions NB was the principal investigator. He operated and followed-up all patients.
FD wrote the protocol, performed the statistical analyses and wrote the study report.
GE, JG, OG teamed with NB to interpret the results clinically. They teamed with NB and FD to write the manuscript.