The main finding from the present study was significantly lower scores for activity-related pain in favor of the cemented group at both two- and twelve-month follow-up. Also, pain at rest at twelve-month follow-up and KOOS-PS at two-month follow-up were significantly lower in the cemented group. Duration of surgery was estimated to be eight minutes less with the cementless technique. A relatively high number of PJIs were found for the cementless group.
The differences in scores between the cemented and the cementless group were less than 1 point, and an important question is whether the statistically significant differences found were clinically significant. Bird and Dickson [7] reported that patients at the upper end of the pain scale usually needed a larger decrease before experiencing pain relief. For the NRS, pain from one to three points is considered mild, whereas, moderate pain ranges from four to six points [10]. The patients in the present study had preoperative mean values around six for activity-related pain and four for pain at rest. Already at the two-month follow-up the NRS was around two in both groups. Bandholm et al. [4] used 1.5 points as the minimal clinically important difference (MID) when evaluating knee pain during strength training after TKA, while Rian et al. [19] used a MID of 0.5 points when evaluating postoperative pain after TKA. Based on these considerations, the differences between the groups found in the present study could be clinically relevant.
Nam et al. [16] found no significant difference in pain when comparing cemented and cementless TKA four to six-weeks postoperatively. However, they discussed the possibility of increased pain during the early postoperative phase prior to biologic fixation in the cementless group [16]. Radiostereometric analyses have shown increased migration of cementless tibial components compared with cemented components during the first-three months following the operation, before stabilizing subsequently [11, 17]. Another study reported slightly more pain at six months with cementless tibial fixation in TKA compared with cemented fixation, and no difference one-year postoperatively [6]. The differences in pain in the present study, with lower pain levels in the cemented group, were still present after twelve months. Improvements for all PROMs were found for both groups from preoperative score to the twelve-month follow-up evaluation. Kerens et al. [12] found comparable clinical results between their groups when comparing 60 cases of cemented UKA with 60 cases of cementless UKA.
PJI is one of the most severe complications following arthroplasty surgery. For the cementless group, there were a relatively high number of infections during the first twelve months post-operation compared with the cemented group. Overweight and prolonged duration of surgery have been reported to increase risk of infection [20]. There were no differences in preoperative BMI between the two groups in the present study. However, with mean BMI close to 30 kg/m2, a significant number of the patients were overweight (BMI 25–29 kg/m2) or suffered from obesity (BMI > 30 kg/m2). The duration of surgery was, as expected, significantly lower in the cementless group. A change in operative technique or implant could possibly have led to an increase in rate of infection for a period of time. The infections in the present study were evenly distributed within the two time periods studied and could not easily be explained by a learning curve for the new fixation technique. A theoretical advantage for cemented implants, that they allow local antibiotics around the implant from the cement, could have supported the findings in the present study. However, this has not been possible to detect in large registry-based or case–control studies [20]. One patient in the present study underwent osteosynthesis of a fissure near the tibial component and one was in need of a revision due to subsidence of the tibial plateau, both in the cementless group. A recent review found comparable incidence of periprosthetic tibial fractures in cemented and cementless UKA, but the authors discussed elements of the cementless technique that could increase the risk of fracture [8].
A randomized controlled study design could have secured a more random distribution of patients in the two groups, but no large baseline differences were seen between the two groups (Table 1). The follow-up evaluations were performed by an experienced physiotherapist and not by an orthopedic specialist, which could possibly be considered a limitation of the study. This decision was made for logistical reasons and to secure evaluation of an independent observer. Several studies reporting results after UKA have their origins in the environment of the developers of the implants and there have been concerns around the reproducibility of such results [13]. In a review, Labek et al. [13] found in that the institution that developed the Oxford unicompartmental knee replacement was involved in more than 50% of publications relating to this technique. In the present study, prospectively collected data from a university hospital were reported and three surgeons were involved. This should strengthen the generalizability of the study. Also, the demographic data correspond well to other publications summarized in a recent review [5]. Together with a follow-up rate well above 90%, the presented results should be very representative of this patient group and thus should be very useful to others. The PROMs in the present study, together with the relatively high number of PJIs in the cementless group, were sufficient to question further use of cementless UKA. The department, therefore, has returned to using the cemented technique for all UKAs.