What are the age-related factors linked to aseptic revisions in constrained and unconstrained TKA as well as UKA? A register-based study from the German arthroplasty registry (EPRD)

Aseptic reasons are a major cause of surgical revision after primary knee arthroplasty [14, 15]. Aseptic revision rates vary throughout researches and procedures [9, 15‐17]. An analysis of the Danish knee arthroplasty register from 1997 to 2017 reported higher revision risk and lower mortality risk for UKA vs. TKA at all time points [18]. When comparing TKA patients with UKA patients, the revision risk for the majority of present patients dropped over the previous 20 years from a 3-year HR of approximately 5 to an HR of 1.5 [18]. Moreover, revision surgery is anticipated to be necessary for a higher percentage of younger individuals as the number of younger patients undergoing TKA rises [8]. Young patients at the time of primary TKA have been linked to increased rates of reoperation and failure [19]. Mc Calden et al. investigated the patient outcome and the revision rate in different age groups in TKA, however no subdivision into unconstrained and constrained TKA was carried out. For this purpose, the cohort was divided into three age groups: patients younger than 55 years, patients between 55 and 70 years and patients older than 70 years. A higher revision rate was found in patients younger than 55 years. The overall revision rate was 5.9% for patients younger than 55 years, 3.1% for patients between 55 and 70 years and 2.1% for patients older than 70 years [11]. Charette et al. also reported a higher rate of revision in young patients [20]. In the present analysis the cohort was divided into patients younger than 65 years, patients aged between 65 and 74 years and patients older than 74 years. Our analysis demonstrated comparable results with a significant increased aseptic rate of revision in patients younger than 65 years compared to patients aged between 65 and 74 years and patients older than 75 years (p < 0.0001). Moreover, young patients under 55 years were more likely than patients older than 55 years to undergo early revision of TKA within two years following primary TKA (52.5% vs. 29.0%) [21]. Even though TKA is an excellent therapy option for younger patients with osteoarthritis of the knee, it is crucial to inform these patients about the lower survival rate associated with knee arthroplasty. Compared to older patient groups, younger patients had poorer TKA survivability but equal or possibly superior clinical results after primary TKA [11]. Early-life history of knee prosthesis implantation is associated with a higher likelihood of co-morbidities and potentially poorer health characteristics [5, 19, 21]. According to Keeney et al. a higher body mass index and lower activity levels are found among younger patients undergoing primary knee arthroplasty [22]. One the one hand, this leads to a higher revision rate due to septic reasons because of an increased risk for infection. On the other hand, an increased risk of aseptic mechanical failure due to the patients’ weight could be observed [13, 26]. Moreover, in younger patients, sustained high activity levels are unlikely to be a main cause for aseptic revision of primary knee arthroplasties [22]. Comparable results could be achieved in the case of a UKA. The risk of revision was once more highest in the youngest age group from 46 to 50 years with 40.4% and decreased sequentially until the oldest age group from 86 to 90 years which accounted 3.7% [23]. Younger patients have a higher risk of revision due to progression of osteoarthritis of the knee or aseptic loosening [24, 25]. In UKA, younger age is thought to correlate with increased activity, which may accelerate wear, aseptic loosening or progression of osteoarthritis [26]. In addition, younger patients may have greater expectations of their post-operative function and activity level. Therefore, further surgery will be discussed if pain returns [26].

In the present investigation, the risk and preventive factors for UKA as well as constrained and unconstrained TKA were analyzed according to age groups.

In constrained TKA, neither weight, gender, Elixhauser score or TKA volume could be identified as a risk or preventive factor in all three age groups. Studies examining influencing factors in constrained and unconstrained TKA as well as UKA with regard to the individual age groups are rare. An analysis by the New Zealand Joint Registry from January 1999 to Dezember 2016 for risk factors of TKA revision demonstrated that patients between 45 and 50 years of age have the worst 10-year implant survival rate of 91.5%, and the implantat survival rate gradually improved to 99.2% in patients between 90 and 95 years of age [27]. In our study a time split after six months was conducted to reduce the violation of confounding variables when assuming constant proportional risks.

In unconstrained TKA, an Elixhauser score of 1–4 (HR = 1,7; p < 0,001) and an Elixhauser score ≥ 5 (HR = 1,75; p < 0,001) for patients younger than 65 years as well as an Elixhauser score 1–4 (HR = 1,26;p = 0,025) and an Elixhauser score ≥ 5 (HR = 1,54;p < 0,001) in patients aged between 65 and 74 years were detected as risk factors. Moreover, an Elixhauser score 1–4 (HR = 1,63; p = 0,001) and an Elixhauser score ≥ 5 (HR = 1,7;p < 0,001) in patients older than 74 years were identified as risk factors. An analysis by the New Zealand Joint Registry from January 1999 to Dezember 2016 for risk factors of TKA revision identified the highest risk of revision with 24.5% in patients with ASA grade 3 and 4 aged between 46 and 50 years, compared to 18.9% in ASA grade 1 patients in the same age group (p < 0.001) [27]. In contrast to our study, where male was identified as a preventive factor for patients younger than 65 years (HR = 0,87; p = 0,002) and older than 74 years (HR = 0,88; p = 0,018), the analysis of the New Zealand Joint register detected male as a risk factor for revision in TKA [27]. Mikkelsen et al. elucidated, that cementless TKAs had higher revision risks than cemented TKAs (HR 1.7, CI 1.4–1.9) [18]. Moreover, Fleischman et al. reported significantly reduced risk for mechanical failure with cemented TKA by 58.9% (p = 0,0002) with increasing age for each additional decade of life [28]. In the present study, the use of a cemented system in unconstrained TKA (HR = 0,52; p = 0,002) was also identified as a preventive factor in patients older than 74 years. An analyses of Nordic Arthroplasty Register Association between 2000 and 2016 identified uncemented TKA with increased risk of revision compared with the cemented TKA (HR = 1.3) [29]. Namba et al. identified for TKA afro-american patients (HR = 1.73), diabetes (HR = 1.21), a volume of performed TKA under 50 (HR = 1.11), as well as again cementless knee arthroplasty (HR = 1.28) as risk factors for aseptic revision [15]. Gelderman et al. reported post-operative dissatisfaction being more common in younger patients (< 55 years) and younger patients having the double risk to undergo early revision than older patients (> 60–75 years) [30]. However, in literature patient-related preventive factors according to the patients’ age and type of knee prothesis are rarely discussed. Namba et al. mentioned a 37% decreased risk of revision by same day bilateral procedures in TKA [15]. Moreover, an BMI > 35 kg/m² vs. BMI < 30 kg/m² had a lower risk for aseptic revision (HR = 0.78) [15]. In comparison, in the present study, pre-obese (HR = 0,86; p = 0,048), obesity grade I (HR = 0,87; p = 0,046) and obesity grade III (HR = 0,77; p = 0,005) could be identified as preventive factor after six months postoperative in patients under 64 years in unconstrained TKA. However, in the patients aged between 64 and 75 years, obesity grade II (HR = 1,47; p = 0,008) and obesity grade III (HR = 2,44; p < 0,001) were reported as risk factors for unconstrained TKA within six months postoperative, whereas pre-obese (HR = 0,78; p = 0,005) and obesity grade I (HR = 0,81; p = 0,02) were identified as preventive factors after 6 months postoperative. In patients older than 74 years, pre-obese (HR = 1,23;p = 0,031), obesity grade I (HR = 1,37; p = 0,004), obesity grade II (HR = 1,59; p = 0,003) and obesity grade III (HR = 2,7; p < 0,001) were detected as risk factors within six months postoperative, whereas after six months postoperative, obesity grade I (HR = 0,74; p = 0,007) could be found as preventive factor. Thus, the weight’s role as a preventive or risk factor in unconstrained TKA varies depending on the period of time and the patient’s age. Boyer et al. mentioned no influence of Diabetes mellitus status (p = 0.9657), BMI (p = 0.4517) or gender (p = 0.88) on survival for aseptic loosening in TKA [31]. Moreover, in the present study, for patients younger than 65 years, a TKA volume > 250 per year (HR = 0,72; p < 0,001) and a TKA volume of 101–250 per year (HR = 0,66; p > 0,001) after six months postoperative. Furthermore, a TKA volume > 250 per year (HR = 0,5; p < 0,001) within six months postoperative were detected as preventive factors for an aseptic revision in unconstrained TKA. For patients aged between 64 and 75 years, a TKA volume of 101–250 per year (HR = 0,62; p < 0,001) and a TKA volume > 250 per year (HR = 0,69; p < 0,001) within six months postoperative as well as a TKA volume of > 250 per year (HR = 0,78; p = 0,001) after six months postoperative were also identified as preventive factors. For patients older than 74 years, a TKA volume of 101–250 per year (HR = 0,79; p = 0,014) was recorded as preventive factors within six months postoperative. Moreover, Badaway et al. reported increasingly better results with increasing annual hospital volume [32].

In case of UKA, cemented prothesis were analyzed as preventive factor in patients aged between 65 and 74 years (HR = 0,37; p < 0,001) and in patients older than 74 years (HR = 0,37; p < 0,001) within six months postoperative. The identification of patient comorbidities and prosthesis specific influencing factors for aseptic revision of UKA varies among the literature. However, in literature patient-related and prothesis specific influencing factors with regard to the patient age are rarely discussed. Mikkelsen et al. elucidated, that revision risk was lower for cementless UKA compared with cemented UKA (HR 0.6, CI 0.5–78) [18]. Moreover, Tay et al. reported a higher risk of aseptic revision for cemented mobile-bearing UKA compared with cemented fixed-bearing UKA (HR = 1.9, 95% CI = 1.1–3.2; p = 0.03) [33]. Additionally, in our investigation, a UKA volume > 50 per year (HR = 0,62; p < 0,001) in patients younger than 65 years or patients aged between 65 and 74 years (HR = 0,59; p = 0,003) were preventive factors. In comparison, an analyses of the Dutch arthroplasty register from 2007 to 2016, hospitals with higher TKA volume were more likely to use UKA and hospitals with a higher absolute or proportional UKA volume could improve survival [34]. Hospitals with an absolute volume of performed UKA between 22 and 36 per year had a higher risk for revision (HR = 1.04), whereas a UKA volume between 36 and 58 annually (HR = 0.96) and a UKA volume more than 58 (HR = 0.74) annually leaded to a lower risk of aseptic revision [35]. This is comparable to our findings that a high volume of UKA implantations can lead to a satisfactory functional outcome. Younger individuals had a higher likelihood of revision than older patients (≥ 70 years). In UKA, patients under 60 years old demonstrated a 1.9-fold increase in aseptic revision risk, while those between 60 and 69 years old presented a 1.6-fold rise in aseptic revision risk (< 60 years: HR = 1.9, 95% CI = 1.2-3.0; 60–69 years: HR = 1.6, 95% CI = 1.0-2.4; p < 0.05) [33]. Additionally, a higher cumulative frequency of revision for aseptic loosening in UKA in younger patients was observed (3.2% vs. 2.7% for ≥ 70 years; p < 0.05) [33]. Moreover, in the present analyses in all three age categories, male was found as a preventative factor (< 65: HR = 0,81; p < 0,001; 65–74: HR = 0,72; p < 0,00; >74: HR = 0,57; p < 0,001). Additionally, pre-obese (HR = 0,76; p = 0,003), an Elixhauser score 0 (HR = 0,86; p = 0,029) and an Elixhauser score 1–4 (HR = 0,77; p = 0,043) could be identified as a preventive factors for aseptic revision of UKA in patients younger than 65 years. The identification of patient co-morbidities as risk factors for aseptic revision of UKA varies among the literature. However, in literature patient-related influencing factors with regard to the patient age are rarely discussed. Tay et al. reported a 1.2- to 1.5-fold higher lifetime risk of revision for UKA for females compared to males across all age groups. Women have a lower mortality rate than men in all age groups. Therefore, it can be assumed to live longer and be exposed to a higher revision rates [23, 36]. Boyer et al. mentioned no influence of Diabetes status (p = 0.1186) or BMI (p = 0.4561) on survival for aseptic loosening in UKA. However, in contrast to the present investigation, a substantial increase in risk in UKA was discovered for females (p = 0.0175) [31].

Despite multiple advantages of the German Arthroplasty Registry, several limitations of the present study are worth to be mentioned. The accurate coding of procedures and surgeon registration are essential for the quality of data in this registry. The included patient data were cross-validated using insurance data in order to reduce this impact and constraint. Due to different indications of the investigated implants different quantities were reported. Kaplan-Meier-estimates, the corrected multiple Log-Rank-Test as well as a time split after six months for the calculation of hazard ratio were used to reduce this limitation. An additional limitation pertains to the length of the registry’s existence, as it presently precludes the investigation of follow-ups lasting more than seven years. Inaccurate or inadequate coding is another potential cofounder. The Elixhauser Comorbidity score was computed using the comorbidities recorded during the initial hospital stay after primary implantation. However, not all comorbidities are evaluated using the Elixhauser Comorbidity Index. Moreover, the present analysis’s informative value is diminished by the incomprehensibility of the stem lengths for shaft-anchored prosthesis based on registry data. Another limitation is the fact that the degree of coupling cannot be detailed for constraint prostheses. An additional constraint pertains to the analysis of the individual revisions’ causes solely to the type of care, without considering the varying age groups included within that category.